2 x Intel Xeon 6780E testing with a Quanta Cloud QuantaGrid D55Q-2U S7Q-MB-MPS-MDP (2A05.QCT001 BIOS) and ASPEED on Ubuntu 24.04 via the Phoronix Test Suite.
Xeon Platinum 8490H Processor: Intel Xeon Platinum 8490H @ 3.50GHz (60 Cores / 120 Threads), Motherboard: Quanta Cloud QuantaGrid D54Q-2U S6Q-MB-MPS (3B05.TEL4P1 BIOS), Chipset: Intel Device 1bce, Memory: 512GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b000590Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Xeon Platinum 8490H 2P Changed Processor to 2 x Intel Xeon Platinum 8490H @ 3.50GHz (120 Cores / 240 Threads) .
Changed Memory to 1008GB .
Xeon Platinum 8592+ Changed Processor to INTEL XEON PLATINUM 8592+ @ 3.90GHz (64 Cores / 128 Threads) .
Changed Memory to 512GB .
Processor Change: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000200
Xeon Platinum 8592+ 2P Changed Processor to 2 x INTEL XEON PLATINUM 8592+ @ 3.90GHz (128 Cores / 256 Threads) .
Changed Memory to 1008GB .
Xeon Max 9468 Changed Processor to Intel Xeon Max 9468 @ 3.50GHz (48 Cores / 96 Threads) .
Changed Memory to 576GB .
Processor Change: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2c000290
Xeon Max 9468 2P Changed Processor to 2 x Intel Xeon Max 9468 @ 3.50GHz (96 Cores / 192 Threads) .
Changed Memory to 1136GB .
Xeon Max 9480 Changed Processor to Intel Xeon Max 9480 @ 3.50GHz (56 Cores / 112 Threads) .
Changed Memory to 576GB .
Xeon Max 9480 2P Changed Processor to 2 x Intel Xeon Max 9480 @ 3.50GHz (112 Cores / 224 Threads) .
Changed Memory to 1136GB .
Xeon Platinum 8380 2P Processor: 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads)Motherboard: Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS)Chipset: Intel Ice Lake IEHMemory: 512GBDisk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Monitor: VE228Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1080
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd0003d1Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Mitigation of Microcode + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Xeon Platinum 8380 Changed Processor to Intel Xeon Platinum 8380 @ 3.40GHz (40 Cores / 80 Threads) .
Changed Memory to 256GB .
EPYC 9684X 2P Processor: 2 x AMD EPYC 9684X 96-Core @ 2.55GHz (192 Cores / 384 Threads)Motherboard: AMD Titanite_4G (RTI1007B BIOS)Chipset: AMD Device 14a4Memory: 1520GBDisk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: Broadcom NetXtreme BCM5720 PCIe
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa101144Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
EPYC 9684X Changed Processor to AMD EPYC 9684X 96-Core @ 2.55GHz (96 Cores / 192 Threads) .
Changed Memory to 12 x 64GB DDR5-4800MT/s Samsung M321R8GA0BB0-CQKEG .
EPYC 9654 Changed Processor to AMD EPYC 9654 96-Core @ 2.40GHz (96 Cores / 192 Threads) .
EPYC 9654 2P Changed Processor to 2 x AMD EPYC 9654 96-Core @ 2.40GHz (192 Cores / 384 Threads) .
Changed Memory to 1520GB .
EPYC 9754 Changed Processor to AMD EPYC 9754 128-Core @ 2.25GHz (128 Cores / 256 Threads) .
Changed Memory to 12 x 64GB DDR5-4800MT/s Samsung M321R8GA0BB0-CQKEG .
Processor Change: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa00116
EPYC 9754 2P Changed Processor to 2 x AMD EPYC 9754 128-Core @ 2.25GHz (256 Cores / 512 Threads) .
Changed Memory to 1520GB .
EPYC 8534P Processor: AMD EPYC 8534P 64-Core @ 2.30GHz (64 Cores / 128 Threads)Motherboard: AMD Cinnabar (RCB1009C BIOS)Chipset: AMD Device 14a4, Memory: 6 x 32GB DDR5-4800MT/s Samsung M321R4GA0BB0-CQKMGDisk: 2 x 1920GB KIOXIA KCD8XPUG1T92 + 2000GB Corsair MP700 PROGraphics: ASPEED, Network: 2 x Broadcom NetXtreme BCM5720 PCIe
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa00213Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
EPYC 8534PN Changed Processor to AMD EPYC 8534PN 64-Core @ 2.00GHz (64 Cores / 128 Threads) .
Changed Disk to 2 x 1920GB KIOXIA KCD8XPUG1T92 .
Xeon 6766E 2P Processor: 2 x Intel Xeon 6766E @ 2.70GHz (288 Cores)Motherboard: Quanta Cloud QuantaGrid D55Q-2U S7Q-MB-MPS-MDP (2A05.QCT001 BIOS)Chipset: Intel Ice Lake IEHMemory: 512GBDisk: 3201GB Micron_7450_MTFDKCB3T2TFS + 2 x 1920GB KIOXIA KCD8XPUG1T92Graphics: ASPEED, Network: 2 x Broadcom BCM57416 NetXtreme-E Dual-Media 10G RDMA
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x13000131Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Xeon 6766E Changed Processor to Intel Xeon 6766E @ 2.70GHz (144 Cores) .
Changed Memory to 256GB .
Changed Disk to 2 x 1920GB KIOXIA KCD8XPUG1T92 + 3201GB Micron_7450_MTFDKCB3T2TFS .
Xeon 6780E Changed Processor to Intel Xeon 6780E @ 3.00GHz (144 Cores) .
Changed Disk to 3201GB Micron_7450_MTFDKCB3T2TFS + 2 x 1920GB KIOXIA KCD8XPUG1T92 .
Xeon 6780E 2P Changed Processor to 2 x Intel Xeon 6780E @ 3.00GHz (288 Cores) .
Changed Memory to 512GB .
WRF WRF, the Weather Research and Forecasting Model, is a "next-generation mesoscale numerical weather prediction system designed for both atmospheric research and operational forecasting applications. It features two dynamical cores, a data assimilation system, and a software architecture supporting parallel computation and system extensibility." Learn more via the OpenBenchmarking.org test page.
OpenVKL OpenVKL is the Intel Open Volume Kernel Library that offers high-performance volume computation kernels and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Items / Sec, More Is Better OpenVKL 2.0.0 Benchmark: vklBenchmarkCPU ISPC EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9754 EPYC 9684X Xeon 6780E 2P EPYC 9654 Xeon 6766E 2P Xeon Platinum 8592+ Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon 6780E Xeon Max 9480 Xeon 6766E EPYC 8534P Xeon Max 9468 EPYC 8534PN Xeon Platinum 8380 900 1800 2700 3600 4500 SE +/- 26.33, N = 3 SE +/- 5.36, N = 3 SE +/- 8.67, N = 3 SE +/- 13.62, N = 3 SE +/- 6.74, N = 3 SE +/- 0.88, N = 3 SE +/- 1.00, N = 3 SE +/- 3.61, N = 3 SE +/- 1.20, N = 3 SE +/- 7.31, N = 3 SE +/- 2.08, N = 3 SE +/- 4.36, N = 3 SE +/- 8.62, N = 3 SE +/- 1.53, N = 3 SE +/- 12.02, N = 3 SE +/- 0.67, N = 3 SE +/- 13.96, N = 3 SE +/- 0.58, N = 3 SE +/- 1.20, N = 3 SE +/- 6.66, N = 3 SE +/- 8.17, N = 3 SE +/- 6.81, N = 3 4012 3906 3335 3222 2716 2508 2396 2228 2077 2020 2004 1852 1776 1734 1527 1391 1356 1267 1266 1243 1177 869 Items / Sec Per Watt
OpenBenchmarking.org Items / Sec Per Watt, More Is Better OpenVKL 2.0.0 Benchmark: vklBenchmarkCPU ISPC EPYC 8534PN EPYC 9754 EPYC 8534P EPYC 9754 2P EPYC 9654 Xeon 6766E EPYC 9684X Xeon Platinum 8592+ Xeon 6780E EPYC 9654 2P Xeon 6766E 2P EPYC 9684X 2P Xeon 6780E 2P Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Max 9468 2P 3 6 9 12 15 9.780 9.457 9.163 7.673 7.601 7.196 6.861 6.261 6.233 6.102 5.786 5.763 5.416 5.391 5.106 4.287 4.129 3.798 3.527 3.162 2.853 2.745 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenVKL 2.0.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 18.05 / Avg: 120.35 / Max: 144.9 Min: 16.98 / Avg: 138.16 / Max: 167.46 Min: 43.67 / Avg: 176.07 / Max: 237.02 Min: 43.61 / Avg: 223.17 / Max: 308.98 Min: 51.77 / Avg: 246.36 / Max: 292 Min: 21.69 / Avg: 265.2 / Max: 336.16 Min: 28.49 / Avg: 273.27 / Max: 332.24 Min: 35.93 / Avg: 320.1 / Max: 378.34 Min: 57.62 / Avg: 321.66 / Max: 378.49 Min: 75.39 / Avg: 327.25 / Max: 376.15 Min: 78.05 / Avg: 328.39 / Max: 378.11 Min: 106.2 / Avg: 349.12 / Max: 504.22 Min: 33.91 / Avg: 349.21 / Max: 411.78 Min: 63.56 / Avg: 411.4 / Max: 636.51 Min: 102.24 / Avg: 482.91 / Max: 579.87 Min: 40.48 / Avg: 522.89 / Max: 678.27 Min: 57.11 / Avg: 546.55 / Max: 678.78 Min: 79.05 / Avg: 630.97 / Max: 751.76 Min: 172.02 / Avg: 633.5 / Max: 755.3 Min: 209.11 / Avg: 647.03 / Max: 756.03 Min: 190.07 / Avg: 649.08 / Max: 754.84 Min: 67.14 / Avg: 677.76 / Max: 821.93
RELION RELION - REgularised LIkelihood OptimisatioN - is a stand-alone computer program for Maximum A Posteriori refinement of (multiple) 3D reconstructions or 2D class averages in cryo-electron microscopy (cryo-EM). It is developed in the research group of Sjors Scheres at the MRC Laboratory of Molecular Biology. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better RELION 4.0.1 Test: Basic - Device: CPU EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9754 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ EPYC 9754 Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 150 300 450 600 750 SE +/- 1.27, N = 15 SE +/- 2.11, N = 12 SE +/- 1.41, N = 3 SE +/- 2.17, N = 3 SE +/- 4.04, N = 12 SE +/- 2.66, N = 12 SE +/- 2.65, N = 3 SE +/- 3.24, N = 12 SE +/- 2.95, N = 3 SE +/- 2.68, N = 7 SE +/- 7.69, N = 9 SE +/- 0.35, N = 3 SE +/- 7.14, N = 9 SE +/- 4.65, N = 9 SE +/- 2.98, N = 3 SE +/- 4.63, N = 3 SE +/- 6.20, N = 3 SE +/- 7.86, N = 9 93.17 145.65 155.74 176.42 201.27 201.38 203.41 214.42 238.88 290.45 360.21 368.22 373.18 384.38 412.37 468.69 496.24 705.58 1. (CXX) g++ options: -fopenmp -std=c++11 -O3 -rdynamic -ldl -ltiff -lfftw3f -lfftw3 -lpng -ljpeg -lmpi_cxx -lmpi CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better RELION 4.0.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.41 / Avg: 148.71 / Max: 151.06 Min: 18.83 / Avg: 163.61 / Max: 166.72 Min: 53.81 / Avg: 267.6 / Max: 290.36 Min: 21.86 / Avg: 270.75 / Max: 336.67 Min: 28.15 / Avg: 320.59 / Max: 347.29 Min: 22.57 / Avg: 341.82 / Max: 377.61 Min: 50.74 / Avg: 344.24 / Max: 377.4 Min: 73.36 / Avg: 344.98 / Max: 377.25 Min: 73.55 / Avg: 345.11 / Max: 376.88 Min: 32.42 / Avg: 394.87 / Max: 409.71 Min: 103.02 / Avg: 531.46 / Max: 582.55 Min: 42.44 / Avg: 553.69 / Max: 696.76 Min: 52.45 / Avg: 622.74 / Max: 703.72 Min: 77.4 / Avg: 665.17 / Max: 755.23 Min: 170.14 / Avg: 679.44 / Max: 754.54 Min: 195.56 / Avg: 682.36 / Max: 754.93 Min: 199.08 / Avg: 683.82 / Max: 754.9 Min: 63.79 / Avg: 752.52 / Max: 819.67
NWChem NWChem is an open-source high performance computational chemistry package. Per NWChem's documentation, "NWChem aims to provide its users with computational chemistry tools that are scalable both in their ability to treat large scientific computational chemistry problems efficiently, and in their use of available parallel computing resources from high-performance parallel supercomputers to conventional workstation clusters." Learn more via the OpenBenchmarking.org test page.
High Performance Conjugate Gradient HPCG is the High Performance Conjugate Gradient and is a new scientific benchmark from Sandia National Lans focused for super-computer testing with modern real-world workloads compared to HPCC. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org GFLOP/s, More Is Better High Performance Conjugate Gradient 3.1 X Y Z: 144 144 144 - RT: 60 Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9654 EPYC 9684X Xeon Max 9468 EPYC 9754 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 15 30 45 60 75 SE +/- 0.94, N = 3 SE +/- 0.07, N = 3 SE +/- 0.03, N = 3 SE +/- 0.29, N = 3 SE +/- 0.51, N = 3 SE +/- 0.05, N = 3 SE +/- 0.02, N = 3 SE +/- 0.08, N = 3 SE +/- 0.02, N = 3 SE +/- 0.37, N = 3 SE +/- 0.34, N = 9 SE +/- 0.05, N = 3 SE +/- 0.28, N = 5 SE +/- 0.08, N = 3 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 69.30 62.64 60.85 54.41 44.95 42.52 41.87 35.14 31.07 30.59 30.11 28.32 27.58 25.24 23.63 22.20 22.13 20.41 1. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -lmpi_cxx -lmpi GFLOP/s Per Watt
OpenBenchmarking.org GFLOP/s Per Watt, More Is Better High Performance Conjugate Gradient 3.1 X Y Z: 144 144 144 - RT: 60 EPYC 8534PN EPYC 8534P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9754 Xeon Max 9480 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9468 2P Xeon Platinum 8380 EPYC 9754 2P Xeon Max 9468 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P 0.038 0.076 0.114 0.152 0.19 0.169 0.156 0.104 0.103 0.101 0.099 0.092 0.090 0.089 0.088 0.084 0.079 0.077 0.074 0.073 0.069 0.062 0.057 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better High Performance Conjugate Gradient 3.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 Xeon Platinum 8380 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Max 9480 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 18.07 / Avg: 131.23 / Max: 136.75 Min: 18.73 / Avg: 141.98 / Max: 160.79 Min: 16.99 / Avg: 238.17 / Max: 311.47 Min: 52.91 / Avg: 266.5 / Max: 286.68 Min: 28.57 / Avg: 273.24 / Max: 316.28 Min: 33.89 / Avg: 330.25 / Max: 391.86 Min: 24.94 / Avg: 340.38 / Max: 366.08 Min: 62.11 / Avg: 344.43 / Max: 357.3 Min: 74.31 / Avg: 344.79 / Max: 363.36 Min: 76.3 / Avg: 345.19 / Max: 369.22 Min: 101.77 / Avg: 529.99 / Max: 565.31 Min: 42.67 / Avg: 563.2 / Max: 666.57 Min: 55.46 / Avg: 612.13 / Max: 685.77 Min: 208.84 / Avg: 683.79 / Max: 753.27 Min: 78.99 / Avg: 685.14 / Max: 717.63 Min: 162.34 / Avg: 688.79 / Max: 719.93 Min: 192.7 / Avg: 690.6 / Max: 722.1 Min: 67.43 / Avg: 722.2 / Max: 816.31
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 512 - Model: ResNet-50 Xeon Platinum 8490H 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 9654 Xeon Max 9480 2P EPYC 9684X 2P EPYC 9754 Xeon Platinum 8380 2P Xeon Max 9480 Xeon 6780E 2P EPYC 9684X Xeon Max 9468 Xeon 6766E 2P EPYC 8534P Xeon 6780E EPYC 8534PN Xeon 6766E Xeon Platinum 8380 40 80 120 160 200 SE +/- 1.39, N = 3 SE +/- 0.81, N = 3 SE +/- 0.61, N = 3 SE +/- 1.83, N = 3 SE +/- 0.01, N = 3 SE +/- 0.79, N = 3 SE +/- 0.48, N = 3 SE +/- 0.34, N = 3 SE +/- 1.80, N = 3 SE +/- 0.31, N = 3 SE +/- 1.62, N = 4 SE +/- 0.21, N = 3 SE +/- 0.61, N = 3 SE +/- 1.22, N = 3 SE +/- 1.42, N = 4 SE +/- 1.17, N = 3 SE +/- 0.93, N = 9 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.08, N = 3 174.95 174.93 169.75 161.90 148.45 145.29 140.53 140.01 138.39 137.50 135.17 127.97 125.14 124.60 121.96 117.53 113.42 101.17 97.67 95.63 91.96 91.75 images/sec Per Watt
OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 512 - Model: ResNet-50 EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon 6766E Xeon Platinum 8490H EPYC 9684X Xeon 6780E EPYC 9754 2P Xeon Max 9480 Xeon Platinum 8380 Xeon Max 9468 EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 0.16 0.32 0.48 0.64 0.8 0.711 0.659 0.553 0.487 0.469 0.441 0.433 0.385 0.383 0.366 0.364 0.349 0.343 0.296 0.290 0.258 0.257 0.254 0.246 0.227 0.212 0.202 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.16.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E Xeon Platinum 8380 EPYC 9654 Xeon Platinum 8592+ EPYC 9684X Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9480 Xeon 6766E 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 16.14 / Avg: 134.5 / Max: 142.97 Min: 18.7 / Avg: 153.51 / Max: 165.78 Min: 43.69 / Avg: 208.71 / Max: 235.06 Min: 20.9 / Avg: 244.22 / Max: 295.24 Min: 43.51 / Avg: 255.22 / Max: 299.17 Min: 53.71 / Avg: 263.16 / Max: 276.44 Min: 30.39 / Avg: 287.56 / Max: 321.33 Min: 33.86 / Avg: 299.63 / Max: 330.67 Min: 35.54 / Avg: 316.96 / Max: 353.37 Min: 74.91 / Avg: 342.39 / Max: 362.9 Min: 61.57 / Avg: 342.65 / Max: 363.24 Min: 85.58 / Avg: 343.75 / Max: 366.49 Min: 106.33 / Avg: 441.95 / Max: 500.12 Min: 40.89 / Avg: 478.51 / Max: 540.78 Min: 102.54 / Avg: 504.58 / Max: 542.99 Min: 108.05 / Avg: 506.95 / Max: 588.39 Min: 59.62 / Avg: 547.08 / Max: 589.2 Min: 83.43 / Avg: 585.9 / Max: 658.36 Min: 66.7 / Avg: 606.04 / Max: 658.36 Min: 161.89 / Avg: 678.16 / Max: 719.66 Min: 208.64 / Avg: 684.02 / Max: 721.63 Min: 207.09 / Avg: 684.53 / Max: 724.92
Xcompact3d Incompact3d Xcompact3d Incompact3d is a Fortran-MPI based, finite difference high-performance code for solving the incompressible Navier-Stokes equation and as many as you need scalar transport equations. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 Input: X3D-benchmarking input.i3d EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9684X Xeon Platinum 8592+ Xeon Max 9480 Xeon Platinum 8490H Xeon Max 9468 EPYC 9654 EPYC 9754 Xeon Platinum 8380 EPYC 8534P EPYC 8534PN 120 240 360 480 600 SE +/- 2.20, N = 12 SE +/- 0.37, N = 3 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 SE +/- 0.05, N = 3 SE +/- 3.26, N = 9 SE +/- 9.13, N = 9 SE +/- 1.19, N = 3 SE +/- 3.06, N = 9 SE +/- 2.51, N = 3 SE +/- 0.14, N = 3 SE +/- 3.88, N = 4 SE +/- 1.25, N = 3 SE +/- 4.73, N = 3 SE +/- 5.39, N = 3 SE +/- 0.13, N = 3 SE +/- 0.92, N = 3 SE +/- 0.22, N = 3 167.13 171.38 187.22 189.69 196.81 248.87 253.32 292.03 309.85 323.25 351.64 357.01 407.59 466.55 534.06 558.38 567.29 568.25 1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9480 EPYC 9684X Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.95 / Avg: 129.67 / Max: 136.25 Min: 19.39 / Avg: 140.74 / Max: 150 Min: 52.82 / Avg: 267.65 / Max: 283.3 Min: 11.29 / Avg: 271.32 / Max: 306.5 Min: 30.46 / Avg: 296.94 / Max: 321.61 Min: 23.89 / Avg: 333.9 / Max: 356.36 Min: 73.01 / Avg: 342.5 / Max: 356.35 Min: 59.93 / Avg: 344.75 / Max: 358.18 Min: 80.13 / Avg: 345.35 / Max: 367.23 Min: 33.87 / Avg: 359.91 / Max: 399.74 Min: 101.33 / Avg: 531.6 / Max: 556.64 Min: 38.21 / Avg: 591.17 / Max: 651.55 Min: 56.03 / Avg: 638.9 / Max: 677.63 Min: 77.71 / Avg: 675.14 / Max: 722.78 Min: 158.32 / Avg: 681.46 / Max: 722.98 Min: 203.67 / Avg: 685.84 / Max: 731.99 Min: 201.23 / Avg: 686.09 / Max: 746.29 Min: 66.18 / Avg: 750.96 / Max: 804.8
CloverLeaf CloverLeaf is a Lagrangian-Eulerian hydrodynamics benchmark. This test profile currently makes use of CloverLeaf's OpenMP version. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better CloverLeaf 1.3 Input: clover_bm16 Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon 6766E 2P Xeon 6780E 2P EPYC 9684X Xeon Max 9468 2P EPYC 9654 2P Xeon Max 9480 2P Xeon Platinum 8380 2P Xeon 6780E Xeon 6766E Xeon Platinum 8490H EPYC 9654 EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8592+ 2P EPYC 9754 Xeon Platinum 8592+ EPYC 8534P EPYC 8534PN Xeon Platinum 8380 110 220 330 440 550 SE +/- 0.74, N = 3 SE +/- 0.59, N = 3 SE +/- 1.60, N = 12 SE +/- 2.04, N = 3 SE +/- 2.24, N = 3 SE +/- 1.56, N = 3 SE +/- 0.12, N = 3 SE +/- 2.11, N = 9 SE +/- 3.41, N = 3 SE +/- 0.00, N = 3 SE +/- 0.11, N = 3 SE +/- 0.13, N = 3 SE +/- 3.79, N = 9 SE +/- 1.02, N = 3 SE +/- 1.43, N = 3 SE +/- 2.03, N = 3 SE +/- 2.02, N = 3 SE +/- 4.99, N = 9 SE +/- 0.54, N = 3 SE +/- 0.79, N = 3 SE +/- 1.08, N = 3 SE +/- 1.58, N = 3 214.80 223.54 226.60 236.81 245.82 263.33 266.10 269.20 295.30 313.50 314.88 328.47 334.30 349.67 351.29 357.57 378.05 400.52 410.59 495.25 498.22 512.36 1. (F9X) gfortran options: -O3 -march=native -funroll-loops -fopenmp CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better CloverLeaf 1.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9684X Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9480 EPYC 9754 2P Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P 130 260 390 520 650 Min: 18.03 / Avg: 132.66 / Max: 136.36 Min: 18.26 / Avg: 143.82 / Max: 147.88 Min: 43.58 / Avg: 186.46 / Max: 191.98 Min: 19.15 / Avg: 200.7 / Max: 237.69 Min: 43.52 / Avg: 205.96 / Max: 212.24 Min: 28.5 / Avg: 250.17 / Max: 287.09 Min: 48.92 / Avg: 267.62 / Max: 276.86 Min: 33.19 / Avg: 289.13 / Max: 310.83 Min: 34.9 / Avg: 301.76 / Max: 361.01 Min: 71.71 / Avg: 345.92 / Max: 368.64 Min: 57.58 / Avg: 346.05 / Max: 371.34 Min: 75.8 / Avg: 346.15 / Max: 373.49 Min: 42.42 / Avg: 388.45 / Max: 400.3 Min: 105.62 / Avg: 440.67 / Max: 468.45 Min: 107.68 / Avg: 466.63 / Max: 492.37 Min: 57.96 / Avg: 483.81 / Max: 500.79 Min: 76.61 / Avg: 524.03 / Max: 570.93 Min: 105.53 / Avg: 529.36 / Max: 542.31 Min: 66.27 / Avg: 538.72 / Max: 566.53 Min: 198.24 / Avg: 677.8 / Max: 709.86 Min: 200.21 / Avg: 680.25 / Max: 711.02 Min: 171.76 / Avg: 685.99 / Max: 712.37
ASKAP ASKAP is a set of benchmarks from the Australian SKA Pathfinder. The principal ASKAP benchmarks are the Hogbom Clean Benchmark (tHogbomClean) and Convolutional Resamping Benchmark (tConvolve) as well as some previous ASKAP benchmarks being included as well for OpenCL and CUDA execution of tConvolve. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Million Grid Points Per Second, More Is Better ASKAP 1.0 Test: tConvolve MT - Gridding EPYC 9684X EPYC 9684X 2P EPYC 9754 EPYC 9654 2P EPYC 9654 EPYC 9754 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Max 9480 2P Xeon Max 9480 Xeon Max 9468 Xeon Max 9468 2P 3K 6K 9K 12K 15K SE +/- 262.30, N = 12 SE +/- 204.28, N = 3 SE +/- 2.09, N = 3 SE +/- 141.43, N = 5 SE +/- 59.24, N = 3 SE +/- 21.01, N = 3 SE +/- 4.15, N = 3 SE +/- 75.19, N = 5 SE +/- 60.72, N = 3 SE +/- 0.88, N = 3 SE +/- 0.88, N = 3 SE +/- 19.36, N = 3 SE +/- 6.18, N = 3 SE +/- 54.27, N = 5 SE +/- 195.28, N = 8 SE +/- 32.87, N = 12 SE +/- 87.89, N = 12 SE +/- 23.56, N = 3 14955.60 14185.10 12731.80 12708.50 12394.60 11208.70 8088.11 7630.06 6038.92 5826.44 5821.46 5279.07 5172.96 5062.07 4382.97 3406.96 3375.32 3347.93 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
Result
OpenBenchmarking.org Million Grid Points Per Second, More Is Better ASKAP 1.0 Test: tConvolve MT - Degridding EPYC 9684X 2P EPYC 9654 EPYC 9684X EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9754 Xeon Platinum 8380 EPYC 8534PN EPYC 8534P 4K 8K 12K 16K 20K SE +/- 889.21, N = 3 SE +/- 128.32, N = 3 SE +/- 212.00, N = 12 SE +/- 345.37, N = 3 SE +/- 267.65, N = 5 SE +/- 322.56, N = 3 SE +/- 25.80, N = 3 SE +/- 3.76, N = 3 SE +/- 0.54, N = 3 SE +/- 34.17, N = 3 16906.20 14217.10 14143.10 13455.00 11941.30 9893.32 9839.64 7425.22 7421.23 7382.43 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp Million Grid Points Per Second Per Watt
OpenBenchmarking.org Million Grid Points Per Second Per Watt, More Is Better ASKAP 1.0 Test: tConvolve MT - Degridding EPYC 9654 EPYC 8534PN EPYC 8534P EPYC 9684X EPYC 9754 EPYC 9754 2P EPYC 9684X 2P Xeon Platinum 8380 EPYC 9654 2P Xeon Platinum 8380 2P 20 40 60 80 100 89.45 85.73 79.95 76.92 76.56 50.18 46.19 40.39 36.68 30.49 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ASKAP 1.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P 120 240 360 480 600 Min: 16.96 / Avg: 86.56 / Max: 131.98 Min: 18.79 / Avg: 92.34 / Max: 148.64 Min: 22.06 / Avg: 128.52 / Max: 260.19 Min: 28.79 / Avg: 158.94 / Max: 298.53 Min: 52.13 / Avg: 183.83 / Max: 281.05 Min: 33.59 / Avg: 183.88 / Max: 366.01 Min: 41.03 / Avg: 268.12 / Max: 506.15 Min: 103.44 / Avg: 324.5 / Max: 557.49 Min: 56.13 / Avg: 325.56 / Max: 537.83 Min: 66.87 / Avg: 365.98 / Max: 703.86
PyTorch This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org batches/sec, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 256 - Model: ResNet-152 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Platinum 8380 2P EPYC 9684X Xeon Platinum 8490H 2P EPYC 9654 Xeon 6780E EPYC 8534P EPYC 8534PN Xeon 6766E EPYC 9754 EPYC 9654 2P EPYC 9684X 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P 6 12 18 24 30 SE +/- 0.29, N = 3 SE +/- 0.08, N = 3 SE +/- 0.15, N = 3 SE +/- 0.15, N = 3 SE +/- 0.05, N = 3 SE +/- 0.09, N = 3 SE +/- 0.16, N = 3 SE +/- 0.18, N = 4 SE +/- 0.08, N = 3 SE +/- 0.20, N = 3 SE +/- 0.13, N = 7 SE +/- 0.11, N = 3 SE +/- 0.19, N = 3 SE +/- 0.06, N = 3 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.15, N = 3 SE +/- 0.13, N = 9 SE +/- 0.13, N = 9 SE +/- 0.04, N = 3 SE +/- 0.06, N = 9 SE +/- 0.12, N = 9 23.33 20.25 19.58 19.20 17.70 16.30 16.00 14.92 14.88 14.08 14.08 13.95 13.64 13.26 13.18 11.88 11.01 7.14 7.02 6.61 6.11 5.71 batches/sec Per Watt
OpenBenchmarking.org batches/sec Per Watt, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 256 - Model: ResNet-152 EPYC 8534PN EPYC 8534P Xeon Platinum 8592+ Xeon 6780E Xeon 6766E Xeon Platinum 8380 Xeon Platinum 8490H EPYC 9654 Xeon Max 9480 Xeon Max 9468 EPYC 9754 EPYC 9684X Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P 0.0284 0.0568 0.0852 0.1136 0.142 0.126 0.114 0.088 0.080 0.080 0.080 0.076 0.068 0.068 0.068 0.067 0.057 0.043 0.030 0.027 0.027 0.025 0.019 0.019 0.018 0.018 0.015 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better PyTorch 2.2.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 110 220 330 440 550 Min: 16.06 / Avg: 104.56 / Max: 112.26 Min: 20.38 / Avg: 115.97 / Max: 125.68 Min: 43.61 / Avg: 148.21 / Max: 158.39 Min: 21.41 / Avg: 165.14 / Max: 176.25 Min: 43.46 / Avg: 170.76 / Max: 186.26 Min: 27.5 / Avg: 205.23 / Max: 220.66 Min: 54.09 / Avg: 220.44 / Max: 241.42 Min: 34.62 / Avg: 245.31 / Max: 264.74 Min: 33.57 / Avg: 265.19 / Max: 292.3 Min: 59.27 / Avg: 266.18 / Max: 295.86 Min: 77.22 / Avg: 282.52 / Max: 309.36 Min: 76.6 / Avg: 288.21 / Max: 321.16 Min: 40.42 / Avg: 321.8 / Max: 358.12 Min: 105.8 / Avg: 329.47 / Max: 356.81 Min: 107.99 / Avg: 344.44 / Max: 394.7 Min: 102.28 / Avg: 349.96 / Max: 424.69 Min: 56.85 / Avg: 395.96 / Max: 433.04 Min: 63.41 / Avg: 457.61 / Max: 501.36 Min: 161.77 / Avg: 525.94 / Max: 589.73 Min: 78.85 / Avg: 540.82 / Max: 576.97 Min: 199.23 / Avg: 598.06 / Max: 629.15 Min: 199.11 / Avg: 604.22 / Max: 644.89
PostgreSQL This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average Latency EPYC 9684X EPYC 9654 EPYC 9684X 2P EPYC 9754 EPYC 9654 2P EPYC 9754 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon 6780E Xeon Max 9468 Xeon Platinum 8490H Xeon 6766E Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Max 9480 Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P 0.2909 0.5818 0.8727 1.1636 1.4545 SE +/- 0.002, N = 3 SE +/- 0.003, N = 3 SE +/- 0.006, N = 12 SE +/- 0.001, N = 3 SE +/- 0.004, N = 12 SE +/- 0.001, N = 3 SE +/- 0.002, N = 3 SE +/- 0.001, N = 3 SE +/- 0.022, N = 9 SE +/- 0.006, N = 5 SE +/- 0.008, N = 12 SE +/- 0.007, N = 3 SE +/- 0.007, N = 3 SE +/- 0.005, N = 3 SE +/- 0.005, N = 12 SE +/- 0.007, N = 3 SE +/- 0.008, N = 12 SE +/- 0.009, N = 12 SE +/- 0.017, N = 9 SE +/- 0.038, N = 9 SE +/- 0.040, N = 12 SE +/- 0.030, N = 12 0.251 0.278 0.289 0.292 0.293 0.329 0.384 0.414 0.525 0.542 0.559 0.572 0.572 0.581 0.589 0.622 0.775 0.815 1.027 1.098 1.149 1.293 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better PostgreSQL 16 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E Xeon Platinum 8380 Xeon 6766E 2P EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon 6780E 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9684X EPYC 9754 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 120 240 360 480 600 Min: 15.23 / Avg: 121.21 / Max: 140.08 Min: 43.55 / Avg: 127.3 / Max: 157.98 Min: 18.63 / Avg: 138.72 / Max: 161.44 Min: 43.45 / Avg: 153.46 / Max: 193.17 Min: 54.25 / Avg: 239.92 / Max: 291.32 Min: 55.77 / Avg: 244.42 / Max: 350.99 Min: 21.35 / Avg: 253.4 / Max: 320.46 Min: 27.65 / Avg: 271.07 / Max: 317.83 Min: 31.51 / Avg: 271.55 / Max: 343.13 Min: 107.71 / Avg: 276.72 / Max: 373.07 Min: 56.77 / Avg: 302.76 / Max: 366.57 Min: 71.87 / Avg: 311.81 / Max: 377.7 Min: 78.53 / Avg: 316.32 / Max: 367.54 Min: 33.83 / Avg: 355.49 / Max: 410.22 Min: 40.49 / Avg: 404.84 / Max: 498.33 Min: 104.48 / Avg: 468.46 / Max: 563.8 Min: 70.21 / Avg: 487.56 / Max: 585.59 Min: 53.13 / Avg: 495.05 / Max: 595.37 Min: 163.81 / Avg: 544.97 / Max: 640.25 Min: 136.51 / Avg: 563.23 / Max: 663.44 Min: 199.74 / Avg: 578.23 / Max: 678.5 Min: 61.31 / Avg: 584.31 / Max: 698.79
OpenBenchmarking.org TPS, More Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Only EPYC 9684X EPYC 9654 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon 6780E Xeon Max 9468 Xeon 6766E Xeon Platinum 8490H Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Max 9480 Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P 900K 1800K 2700K 3600K 4500K SE +/- 27594.68, N = 3 SE +/- 39970.22, N = 3 SE +/- 71870.42, N = 12 SE +/- 41074.86, N = 12 SE +/- 15503.17, N = 3 SE +/- 13473.87, N = 3 SE +/- 11512.67, N = 3 SE +/- 4774.62, N = 3 SE +/- 74539.48, N = 9 SE +/- 19583.16, N = 5 SE +/- 24978.23, N = 12 SE +/- 20221.52, N = 3 SE +/- 20350.34, N = 3 SE +/- 13469.59, N = 3 SE +/- 13863.97, N = 12 SE +/- 17697.00, N = 3 SE +/- 13233.94, N = 12 SE +/- 13874.42, N = 12 SE +/- 16726.04, N = 9 SE +/- 36627.33, N = 9 SE +/- 30933.90, N = 12 SE +/- 18635.21, N = 12 3989358 3603261 3480909 3422731 3422492 3044620 2606500 2416375 1930006 1846080 1792636 1748658 1748146 1720762 1699046 1607424 1292122 1229649 976035 920725 881881 778141 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 256 - Model: ResNet-50 Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9468 2P EPYC 9654 Xeon Max 9480 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ Xeon Platinum 8380 2P EPYC 9754 Xeon Max 9480 EPYC 9684X EPYC 9684X 2P Xeon Max 9468 Xeon 6780E 2P Xeon 6766E 2P EPYC 8534P EPYC 8534PN Xeon 6780E Xeon Platinum 8380 Xeon 6766E 40 80 120 160 200 SE +/- 1.57, N = 3 SE +/- 0.21, N = 3 SE +/- 0.03, N = 3 SE +/- 1.40, N = 3 SE +/- 0.66, N = 3 SE +/- 1.04, N = 12 SE +/- 0.57, N = 3 SE +/- 0.18, N = 3 SE +/- 0.58, N = 3 SE +/- 0.46, N = 3 SE +/- 2.00, N = 12 SE +/- 0.63, N = 3 SE +/- 1.39, N = 4 SE +/- 0.50, N = 3 SE +/- 0.14, N = 3 SE +/- 0.54, N = 3 SE +/- 0.48, N = 3 SE +/- 0.05, N = 3 SE +/- 0.06, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 161.06 150.33 143.67 141.95 138.11 135.28 133.87 131.38 129.02 125.91 125.76 123.69 123.37 120.11 119.73 115.12 103.22 98.70 92.82 92.42 91.65 87.17 images/sec Per Watt
OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 256 - Model: ResNet-50 EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon 6766E Xeon Platinum 8490H EPYC 9684X Xeon 6780E Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 EPYC 9754 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon 6766E 2P Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon Max 9468 2P EPYC 9684X 2P Xeon Max 9480 2P 0.1555 0.311 0.4665 0.622 0.7775 0.691 0.646 0.525 0.485 0.463 0.438 0.425 0.385 0.382 0.363 0.354 0.352 0.291 0.278 0.254 0.252 0.247 0.243 0.242 0.210 0.203 0.200 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.16.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9654 EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 16.01 / Avg: 134.32 / Max: 142.38 Min: 18.72 / Avg: 152.81 / Max: 164.9 Min: 43.62 / Avg: 198.95 / Max: 223.14 Min: 20.95 / Avg: 239.33 / Max: 288.65 Min: 43.44 / Avg: 241.64 / Max: 277.25 Min: 54.97 / Avg: 260.49 / Max: 276.51 Min: 33.15 / Avg: 278.36 / Max: 312.29 Min: 29.3 / Avg: 284.71 / Max: 314.17 Min: 33.82 / Avg: 320.09 / Max: 348.64 Min: 59.06 / Avg: 338.33 / Max: 360.73 Min: 73.46 / Avg: 338.38 / Max: 359.94 Min: 77.09 / Avg: 340.59 / Max: 361.18 Min: 104.82 / Avg: 417.33 / Max: 477.08 Min: 43.33 / Avg: 451.33 / Max: 504.94 Min: 107.7 / Avg: 475.01 / Max: 547.45 Min: 100.88 / Avg: 496.59 / Max: 533.47 Min: 53.87 / Avg: 531.98 / Max: 578.15 Min: 84.37 / Avg: 540.26 / Max: 604.56 Min: 58.73 / Avg: 590.53 / Max: 641.68 Min: 163.23 / Avg: 663.98 / Max: 713.39 Min: 203.91 / Avg: 674.94 / Max: 716.61 Min: 196.92 / Avg: 674.95 / Max: 729.37
OpenSSL OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org verify/s, More Is Better OpenSSL 3.3 Algorithm: RSA4096 EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9754 Xeon 6780E 2P Xeon Max 9480 2P EPYC 9684X Xeon 6766E 2P EPYC 9654 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 8534P Xeon 6780E Xeon Max 9480 EPYC 8534PN Xeon 6766E Xeon Max 9468 Xeon Platinum 8380 800K 1600K 2400K 3200K 4000K SE +/- 2222.55, N = 3 SE +/- 6125.37, N = 3 SE +/- 5.26, N = 3 SE +/- 6745.42, N = 3 SE +/- 862.96, N = 3 SE +/- 444.74, N = 3 SE +/- 205.33, N = 3 SE +/- 277.84, N = 3 SE +/- 1576.96, N = 3 SE +/- 37.76, N = 3 SE +/- 149.48, N = 3 SE +/- 355.69, N = 3 SE +/- 40228.60, N = 3 SE +/- 92.62, N = 3 SE +/- 92.46, N = 3 SE +/- 126.84, N = 3 SE +/- 23.84, N = 3 SE +/- 200.66, N = 3 SE +/- 136.07, N = 3 SE +/- 9.55, N = 3 SE +/- 165.80, N = 3 SE +/- 198.68, N = 3 3729470.8 3214376.3 2861959.8 2057021.6 1916162.9 1872096.1 1644363.7 1627941.4 1613828.7 1479879.0 1436324.9 1395446.4 1099481.5 1036735.6 974189.9 881480.8 826049.5 813667.6 810291.8 743282.7 697707.2 587354.8 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl verify/s Per Watt
OpenBenchmarking.org verify/s Per Watt, More Is Better OpenSSL 3.3 Algorithm: RSA4096 EPYC 8534PN EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 9654 EPYC 9654 2P Xeon 6766E EPYC 9684X EPYC 9684X 2P Xeon 6766E 2P Xeon 6780E Xeon 6780E 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Platinum 8380 Xeon Max 9468 2P Xeon Platinum 8380 2P 1400 2800 4200 5600 7000 6595.71 5836.73 5696.67 5691.68 4919.81 4728.28 4295.14 4121.23 4099.33 4004.98 3735.68 3600.28 3030.66 2997.55 2907.57 2791.77 2489.80 2430.79 2306.66 2215.62 2215.41 2079.56 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9654 Xeon Max 9468 EPYC 9754 Xeon Max 9480 Xeon Platinum 8490H Xeon Platinum 8592+ Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Max 9468 2P EPYC 9754 2P Xeon Max 9480 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.2 / Avg: 122.85 / Max: 135.57 Min: 15.75 / Avg: 154.87 / Max: 162.9 Min: 43.61 / Avg: 173.05 / Max: 179.47 Min: 43.31 / Avg: 221.12 / Max: 231.86 Min: 52.16 / Avg: 265.1 / Max: 290.58 Min: 27.6 / Avg: 291.95 / Max: 323.61 Min: 73.89 / Avg: 302.47 / Max: 335.7 Min: 19.93 / Avg: 320.74 / Max: 351.44 Min: 74.71 / Avg: 326.8 / Max: 367.81 Min: 56.91 / Avg: 335.05 / Max: 351.44 Min: 32.63 / Avg: 342.08 / Max: 377.97 Min: 104.34 / Avg: 369.51 / Max: 384.07 Min: 32.96 / Avg: 391.59 / Max: 411.04 Min: 107.48 / Avg: 456.73 / Max: 473.49 Min: 99.67 / Avg: 528.71 / Max: 581.74 Min: 59.69 / Avg: 605.29 / Max: 659.72 Min: 198.5 / Avg: 629.88 / Max: 694.83 Min: 43.66 / Avg: 654.68 / Max: 716.5 Min: 199.02 / Avg: 669.72 / Max: 754.93 Min: 70.64 / Avg: 686.24 / Max: 753.6 Min: 165.7 / Avg: 686.36 / Max: 737.53 Min: 64.61 / Avg: 784.12 / Max: 822.1
OpenBenchmarking.org sign/s, More Is Better OpenSSL 3.3 Algorithm: RSA4096 EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9754 EPYC 9684X Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 8534P Xeon Max 9480 Xeon 6780E 2P Xeon Max 9468 Xeon 6766E 2P EPYC 8534PN Xeon Platinum 8380 Xeon 6780E Xeon 6766E 20K 40K 60K 80K 100K SE +/- 80.37, N = 3 SE +/- 247.60, N = 3 SE +/- 81.38, N = 3 SE +/- 56.68, N = 3 SE +/- 14.34, N = 3 SE +/- 27.42, N = 3 SE +/- 98.71, N = 3 SE +/- 83.09, N = 3 SE +/- 6.92, N = 3 SE +/- 108.70, N = 3 SE +/- 29.11, N = 3 SE +/- 15.15, N = 3 SE +/- 9.84, N = 3 SE +/- 22.17, N = 3 SE +/- 20.13, N = 3 SE +/- 8.20, N = 3 SE +/- 19.45, N = 3 SE +/- 8.25, N = 3 SE +/- 12.56, N = 3 SE +/- 40.37, N = 3 SE +/- 1.63, N = 3 SE +/- 1.37, N = 3 107830.6 98237.0 71942.0 61499.6 54601.1 53966.9 48996.8 48534.8 45531.0 36371.6 35964.9 31594.4 27431.1 25894.6 25441.5 25098.0 22861.0 22587.4 21047.5 18389.3 12610.8 11348.5 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
Timed Linux Kernel Compilation This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Timed Linux Kernel Compilation 6.8 Build: allmodconfig Xeon 6780E 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon 6766E 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E EPYC 9684X EPYC 9654 Xeon Platinum 8592+ EPYC 9754 Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 EPYC 8534P Xeon Max 9468 EPYC 8534PN Xeon Platinum 8380 130 260 390 520 650 SE +/- 0.90, N = 3 SE +/- 0.98, N = 3 SE +/- 0.56, N = 3 SE +/- 0.81, N = 3 SE +/- 0.49, N = 3 SE +/- 1.31, N = 3 SE +/- 0.86, N = 3 SE +/- 0.73, N = 3 SE +/- 0.56, N = 3 SE +/- 0.67, N = 3 SE +/- 0.76, N = 3 SE +/- 0.37, N = 3 SE +/- 0.49, N = 3 SE +/- 0.72, N = 3 SE +/- 1.05, N = 3 SE +/- 0.92, N = 3 SE +/- 0.80, N = 3 SE +/- 0.67, N = 3 SE +/- 0.91, N = 3 SE +/- 0.61, N = 3 SE +/- 0.78, N = 3 SE +/- 0.35, N = 3 174.27 182.00 189.74 192.10 197.63 210.20 218.57 231.03 246.94 251.20 274.86 290.34 311.83 316.90 337.69 356.99 362.45 384.06 406.41 428.39 438.17 612.78 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Timed Linux Kernel Compilation 6.8 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9480 2P Xeon Max 9468 2P 150 300 450 600 750 Min: 16.08 / Avg: 129.6 / Max: 144.82 Min: 16.91 / Avg: 150.79 / Max: 171.19 Min: 43.57 / Avg: 201.59 / Max: 252.67 Min: 43.47 / Avg: 255.14 / Max: 329.17 Min: 50.2 / Avg: 256.23 / Max: 291.69 Min: 19.75 / Avg: 262.8 / Max: 335.9 Min: 27.23 / Avg: 282.28 / Max: 331.17 Min: 25.82 / Avg: 317.64 / Max: 378.37 Min: 56.91 / Avg: 323.36 / Max: 378.1 Min: 81.1 / Avg: 330.36 / Max: 377.46 Min: 70.17 / Avg: 330.93 / Max: 375.87 Min: 104.5 / Avg: 339.12 / Max: 521.39 Min: 34.02 / Avg: 348.42 / Max: 413.49 Min: 106.41 / Avg: 416.56 / Max: 664.2 Min: 40.54 / Avg: 484.57 / Max: 689.98 Min: 99.55 / Avg: 491.75 / Max: 583.83 Min: 54.45 / Avg: 525.29 / Max: 678.52 Min: 85.15 / Avg: 585.62 / Max: 755.31 Min: 162.47 / Avg: 610.89 / Max: 756.51 Min: 63.59 / Avg: 621.64 / Max: 829.42 Min: 185.35 / Avg: 634.18 / Max: 755.16 Min: 200.9 / Avg: 639.43 / Max: 753.88
Stockfish This is a test of Stockfish, an advanced open-source C++11 chess benchmark that can scale up to 1024 CPU threads. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Nodes Per Second, More Is Better Stockfish 16.1 Chess Benchmark EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 EPYC 9754 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E Xeon Platinum 8592+ Xeon Platinum 8380 2P EPYC 8534P Xeon Platinum 8490H EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 90M 180M 270M 360M 450M SE +/- 919073.22, N = 3 SE +/- 2356843.91, N = 3 SE +/- 7961094.38, N = 12 SE +/- 3095674.84, N = 3 SE +/- 2692064.58, N = 4 SE +/- 2482737.87, N = 3 SE +/- 2268830.04, N = 3 SE +/- 3170744.86, N = 15 SE +/- 4652276.19, N = 12 SE +/- 2148606.38, N = 4 SE +/- 2925715.37, N = 11 SE +/- 4344961.90, N = 10 SE +/- 1441726.26, N = 3 SE +/- 3565305.92, N = 15 SE +/- 2966712.90, N = 9 SE +/- 1325249.25, N = 3 SE +/- 2397455.91, N = 12 SE +/- 1304590.97, N = 3 SE +/- 2147688.02, N = 12 SE +/- 629927.17, N = 3 SE +/- 1751369.24, N = 12 SE +/- 1466028.98, N = 9 399124223 370751530 301202232 266919825 235445499 222610378 214524659 197708350 189763799 187761769 149409082 137977203 125194066 122794246 109766345 104597200 99198144 96887479 90658156 73592230 68019088 51099545 1. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fno-peel-loops -fno-tracer -pedantic -O3 -funroll-loops -msse -msse3 -mpopcnt -mavx2 -mbmi -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto-partition=one -flto=jobserver Nodes Per Second Per Watt
OpenBenchmarking.org Nodes Per Second Per Watt, More Is Better Stockfish 16.1 Chess Benchmark EPYC 8534PN EPYC 9754 EPYC 9654 EPYC 8534P EPYC 9684X EPYC 9654 2P EPYC 9684X 2P Xeon 6766E Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P Xeon 6780E Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9480 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Max 9468 Xeon Platinum 8380 130K 260K 390K 520K 650K 621037.11 571475.98 569277.48 568916.36 555213.84 540060.41 520707.98 509643.71 496138.40 470216.10 431231.39 403742.35 331127.60 325017.74 285328.34 276776.39 219573.26 215737.78 202981.45 201379.09 199243.06 193986.52 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Stockfish 16.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9654 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9654 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.19 / Avg: 145.98 / Max: 155.71 Min: 17.26 / Avg: 174.36 / Max: 184.62 Min: 43.78 / Avg: 240.94 / Max: 270.17 Min: 53.05 / Avg: 263.42 / Max: 290.77 Min: 43.61 / Avg: 310.08 / Max: 337.09 Min: 21.51 / Avg: 332.06 / Max: 363.69 Min: 22.35 / Avg: 337.72 / Max: 370.84 Min: 56.83 / Avg: 339.56 / Max: 379.91 Min: 74.29 / Avg: 341.12 / Max: 366.15 Min: 71.31 / Avg: 341.39 / Max: 379.06 Min: 30.05 / Avg: 347.3 / Max: 373.1 Min: 35.5 / Avg: 386.38 / Max: 410.66 Min: 105.79 / Avg: 474.56 / Max: 541.15 Min: 100.78 / Avg: 519.4 / Max: 581.59 Min: 109.43 / Avg: 618.97 / Max: 701.47 Min: 37.25 / Avg: 640.56 / Max: 735.74 Min: 78.55 / Avg: 672.28 / Max: 755.5 Min: 166.29 / Avg: 678.39 / Max: 758.17 Min: 203.83 / Avg: 679.75 / Max: 756.48 Min: 195.85 / Avg: 680.45 / Max: 758.53 Min: 57.4 / Avg: 686.5 / Max: 751.79 Min: 70.77 / Avg: 766.5 / Max: 822.06
Apache IoTDB Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100 EPYC 9684X EPYC 9654 EPYC 9754 EPYC 9654 2P EPYC 9684X 2P EPYC 8534PN EPYC 8534P EPYC 9754 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon 6780E Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6766E Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8380 Xeon 6780E 2P Xeon Platinum 8380 2P Xeon 6766E 2P 20 40 60 80 100 SE +/- 0.23, N = 3 SE +/- 0.31, N = 3 SE +/- 0.10, N = 3 SE +/- 0.65, N = 12 SE +/- 0.59, N = 12 SE +/- 0.03, N = 3 SE +/- 0.31, N = 3 SE +/- 0.79, N = 12 SE +/- 0.68, N = 3 SE +/- 0.80, N = 7 SE +/- 0.15, N = 3 SE +/- 0.15, N = 3 SE +/- 0.87, N = 3 SE +/- 0.48, N = 3 SE +/- 0.11, N = 3 SE +/- 1.18, N = 3 SE +/- 1.01, N = 3 SE +/- 1.11, N = 4 SE +/- 0.43, N = 3 SE +/- 0.65, N = 3 SE +/- 0.53, N = 3 SE +/- 1.37, N = 3 52.93 53.77 58.11 61.07 61.46 64.93 65.66 66.54 81.73 82.73 84.59 85.16 89.53 90.02 90.31 92.23 94.84 95.39 100.29 101.49 102.67 107.28 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Apache IoTDB 1.2 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8380 Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 120 240 360 480 600 Min: 16.39 / Avg: 76.32 / Max: 126.8 Min: 33.17 / Avg: 80.98 / Max: 175.95 Min: 17.66 / Avg: 82.51 / Max: 146.09 Min: 43.43 / Avg: 87.52 / Max: 193.65 Min: 21.8 / Avg: 110.92 / Max: 210.79 Min: 25.56 / Avg: 138.03 / Max: 271.74 Min: 34.95 / Avg: 157 / Max: 310.81 Min: 33.72 / Avg: 171.59 / Max: 347.45 Min: 55.72 / Avg: 176.97 / Max: 287.98 Min: 104.7 / Avg: 186.81 / Max: 319.27 Min: 60.48 / Avg: 192.21 / Max: 328.62 Min: 60.93 / Avg: 216.73 / Max: 347.32 Min: 41.52 / Avg: 220.99 / Max: 367.41 Min: 72.55 / Avg: 249.87 / Max: 361.66 Min: 86.91 / Avg: 254.04 / Max: 360.99 Min: 56.18 / Avg: 264.94 / Max: 455.88 Min: 74.41 / Avg: 298.36 / Max: 505.37 Min: 106.1 / Avg: 305.48 / Max: 479.29 Min: 77 / Avg: 337.01 / Max: 603.62 Min: 171.93 / Avg: 425.12 / Max: 596.94 Min: 212.09 / Avg: 494.07 / Max: 662.92 Min: 218.55 / Avg: 495.29 / Max: 662.8
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100 EPYC 9684X EPYC 9654 EPYC 9754 EPYC 9654 2P EPYC 9684X 2P EPYC 8534PN EPYC 8534P EPYC 9754 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon 6780E Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6766E Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8380 Xeon 6780E 2P Xeon Platinum 8380 2P Xeon 6766E 2P 30M 60M 90M 120M 150M SE +/- 538692.55, N = 3 SE +/- 665173.47, N = 3 SE +/- 263772.09, N = 3 SE +/- 1249999.63, N = 12 SE +/- 1058559.19, N = 12 SE +/- 41021.96, N = 3 SE +/- 576868.09, N = 3 SE +/- 1225528.40, N = 12 SE +/- 739343.51, N = 3 SE +/- 862582.24, N = 7 SE +/- 174748.00, N = 3 SE +/- 276414.35, N = 3 SE +/- 782176.98, N = 3 SE +/- 435900.66, N = 3 SE +/- 84917.07, N = 3 SE +/- 928557.10, N = 3 SE +/- 594234.55, N = 3 SE +/- 872659.60, N = 4 SE +/- 306106.55, N = 3 SE +/- 207721.96, N = 3 SE +/- 365828.34, N = 3 SE +/- 769517.20, N = 3 144760448 142102254 131781309 125595607 124947288 117884977 116879799 115317834 94616807 92771639 90973414 90871770 86354518 85744542 85117897 83651785 80770465 80139566 76729495 75740139 75178480 71664005
PyTorch This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org batches/sec, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 512 - Model: ResNet-152 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P EPYC 9654 EPYC 9684X Xeon 6780E EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 EPYC 9654 2P EPYC 9684X 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P 6 12 18 24 30 SE +/- 0.16, N = 3 SE +/- 0.04, N = 3 SE +/- 0.18, N = 7 SE +/- 0.18, N = 7 SE +/- 0.13, N = 3 SE +/- 0.09, N = 3 SE +/- 0.18, N = 3 SE +/- 0.17, N = 3 SE +/- 0.16, N = 3 SE +/- 0.09, N = 3 SE +/- 0.18, N = 3 SE +/- 0.12, N = 3 SE +/- 0.19, N = 3 SE +/- 0.09, N = 3 SE +/- 0.11, N = 3 SE +/- 0.14, N = 3 SE +/- 0.02, N = 3 SE +/- 0.14, N = 9 SE +/- 0.13, N = 9 SE +/- 0.07, N = 3 SE +/- 0.07, N = 3 SE +/- 0.11, N = 9 23.41 20.46 19.47 19.29 17.55 16.68 16.17 14.76 14.74 14.59 14.05 13.95 13.61 13.28 13.28 11.74 10.91 7.20 7.07 6.65 6.33 5.75 batches/sec Per Watt
OpenBenchmarking.org batches/sec Per Watt, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 512 - Model: ResNet-152 EPYC 8534PN EPYC 8534P Xeon Platinum 8592+ Xeon 6780E Xeon Platinum 8380 Xeon 6766E Xeon Platinum 8490H Xeon Max 9468 EPYC 9654 EPYC 9754 Xeon Max 9480 EPYC 9684X Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P 0.0286 0.0572 0.0858 0.1144 0.143 0.127 0.114 0.088 0.080 0.080 0.079 0.077 0.069 0.068 0.066 0.066 0.057 0.042 0.030 0.029 0.028 0.024 0.019 0.018 0.018 0.018 0.016 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better PyTorch 2.2.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 110 220 330 440 550 Min: 17.18 / Avg: 104.85 / Max: 112.82 Min: 20.79 / Avg: 116.41 / Max: 125.98 Min: 43.61 / Avg: 148.25 / Max: 158.74 Min: 22.02 / Avg: 164.54 / Max: 175.34 Min: 43.48 / Avg: 170.8 / Max: 185.92 Min: 29.11 / Avg: 205.3 / Max: 221.54 Min: 55.66 / Avg: 220.18 / Max: 241.46 Min: 35.28 / Avg: 245.06 / Max: 264.04 Min: 33.85 / Avg: 264.62 / Max: 291.43 Min: 60.34 / Avg: 265.81 / Max: 296.61 Min: 74.15 / Avg: 282.4 / Max: 310.35 Min: 70.87 / Avg: 291.39 / Max: 322.93 Min: 40.81 / Avg: 321.68 / Max: 356.16 Min: 104.97 / Avg: 340.5 / Max: 358.58 Min: 96.51 / Avg: 349.52 / Max: 424.02 Min: 108.01 / Avg: 376.47 / Max: 395.86 Min: 57.69 / Avg: 396.9 / Max: 434.4 Min: 66.98 / Avg: 454.4 / Max: 498.76 Min: 167.37 / Avg: 511.16 / Max: 589.71 Min: 83.37 / Avg: 541.35 / Max: 576.05 Min: 204.68 / Avg: 599.73 / Max: 629.45 Min: 191.23 / Avg: 602.51 / Max: 639.79
Result
OpenBenchmarking.org batches/sec, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 64 - Model: ResNet-152 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P EPYC 9684X EPYC 9654 EPYC 8534P EPYC 8534PN Xeon 6780E Xeon 6766E EPYC 9754 EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P 6 12 18 24 30 SE +/- 0.34, N = 3 SE +/- 0.18, N = 3 SE +/- 0.17, N = 12 SE +/- 0.19, N = 5 SE +/- 0.05, N = 3 SE +/- 0.15, N = 12 SE +/- 0.21, N = 3 SE +/- 0.12, N = 3 SE +/- 0.05, N = 3 SE +/- 0.17, N = 3 SE +/- 0.10, N = 3 SE +/- 0.17, N = 3 SE +/- 0.16, N = 3 SE +/- 0.03, N = 3 SE +/- 0.13, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 SE +/- 0.15, N = 9 SE +/- 0.13, N = 9 SE +/- 0.05, N = 3 SE +/- 0.07, N = 3 SE +/- 0.03, N = 3 23.49 20.13 19.24 19.01 17.73 16.25 15.84 15.11 14.96 14.03 13.89 13.84 13.19 13.15 13.00 11.95 10.88 7.08 7.04 6.59 6.15 5.79 batches/sec Per Watt
OpenBenchmarking.org batches/sec Per Watt, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 64 - Model: ResNet-152 EPYC 8534PN EPYC 8534P Xeon Platinum 8592+ Xeon 6766E Xeon Platinum 8380 Xeon 6780E Xeon Platinum 8490H EPYC 9654 Xeon Max 9468 EPYC 9754 Xeon Max 9480 EPYC 9684X Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P 0.0284 0.0568 0.0852 0.1136 0.142 0.126 0.114 0.089 0.081 0.080 0.076 0.076 0.068 0.067 0.066 0.066 0.057 0.043 0.030 0.027 0.027 0.025 0.019 0.018 0.018 0.018 0.016 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better PyTorch 2.2.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 110 220 330 440 550 Min: 15.84 / Avg: 103.98 / Max: 111.57 Min: 17.98 / Avg: 115.9 / Max: 125.2 Min: 43.61 / Avg: 148.31 / Max: 158.76 Min: 21.08 / Avg: 164.98 / Max: 175.44 Min: 43.44 / Avg: 170.28 / Max: 183.58 Min: 26.78 / Avg: 205 / Max: 221.09 Min: 55.98 / Avg: 220.28 / Max: 241.66 Min: 32.56 / Avg: 244.28 / Max: 263 Min: 34.31 / Avg: 264.8 / Max: 291.45 Min: 59.17 / Avg: 265.52 / Max: 295.66 Min: 73.58 / Avg: 282.03 / Max: 308.67 Min: 77.61 / Avg: 291.69 / Max: 322.84 Min: 40.59 / Avg: 320.17 / Max: 358.77 Min: 104.9 / Avg: 330.33 / Max: 356.82 Min: 100.57 / Avg: 349.16 / Max: 423.16 Min: 107.82 / Avg: 374.58 / Max: 394.46 Min: 39.14 / Avg: 396.83 / Max: 432.36 Min: 64.18 / Avg: 455.72 / Max: 501.58 Min: 161.59 / Avg: 520.21 / Max: 588.55 Min: 76.4 / Avg: 541.7 / Max: 581.39 Min: 211.71 / Avg: 595.73 / Max: 626.88 Min: 193.13 / Avg: 605.19 / Max: 641.47
SecureMark SecureMark is an objective, standardized benchmarking framework for measuring the efficiency of cryptographic processing solutions developed by EEMBC. SecureMark-TLS is benchmarking Transport Layer Security performance with a focus on IoT/edge computing. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org marks, More Is Better SecureMark 1.0.4 Benchmark: SecureMark-TLS Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9684X 2P EPYC 9684X EPYC 9654 2P EPYC 9654 Xeon Max 9468 Xeon Max 9468 2P Xeon Max 9480 2P Xeon Max 9480 Xeon Platinum 8490H Xeon Platinum 8490H 2P EPYC 9754 2P EPYC 9754 EPYC 8534PN EPYC 8534P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E 2P Xeon 6766E 70K 140K 210K 280K 350K SE +/- 209.97, N = 3 SE +/- 411.11, N = 3 SE +/- 809.90, N = 3 SE +/- 415.71, N = 3 SE +/- 389.81, N = 3 SE +/- 361.20, N = 3 SE +/- 256.85, N = 3 SE +/- 384.08, N = 3 SE +/- 196.24, N = 3 SE +/- 670.43, N = 3 SE +/- 625.56, N = 3 SE +/- 330.95, N = 3 SE +/- 553.62, N = 3 SE +/- 457.42, N = 3 SE +/- 542.88, N = 3 SE +/- 372.96, N = 3 SE +/- 372.88, N = 3 SE +/- 285.52, N = 3 SE +/- 406.59, N = 3 SE +/- 1219.44, N = 3 SE +/- 722.69, N = 3 SE +/- 553.57, N = 3 309896 308434 301740 300833 299610 298341 277647 277589 277298 277260 277205 275472 251075 250695 250162 250024 231357 231076 220051 217536 197298 196740 1. (CC) gcc options: -pedantic -O3 marks Per Watt
OpenBenchmarking.org marks Per Watt, More Is Better SecureMark 1.0.4 Benchmark: SecureMark-TLS Xeon 6780E EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8592+ EPYC 9754 EPYC 9654 Xeon Platinum 8490H EPYC 9684X Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 1100 2200 3300 4400 5500 4990.15 4781.86 4697.25 4501.16 3643.38 3204.23 3034.03 3029.20 2686.16 2364.52 2308.76 2198.71 2010.97 1847.99 1278.59 1272.88 1242.56 1197.70 1185.32 1152.38 986.37 973.55 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better SecureMark 1.0.4 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 EPYC 9654 Xeon 6766E 2P Xeon 6780E 2P EPYC 9684X Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9754 2P Xeon Platinum 8490H 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 60 120 180 240 300 Min: 43.5 / Avg: 43.71 / Max: 45.5 Min: 43.33 / Avg: 44.1 / Max: 46.18 Min: 14.73 / Avg: 52.31 / Max: 55.82 Min: 16.16 / Avg: 53.23 / Max: 56.82 Min: 18.81 / Avg: 78.24 / Max: 81.48 Min: 33.51 / Avg: 85.06 / Max: 96.45 Min: 53.54 / Avg: 91.51 / Max: 104.77 Min: 49.93 / Avg: 97.85 / Max: 101.93 Min: 28.33 / Avg: 98.33 / Max: 101.2 Min: 103.87 / Avg: 106.76 / Max: 110.05 Min: 106.5 / Avg: 108.17 / Max: 111.72 Min: 31.06 / Avg: 111.99 / Max: 116.51 Min: 74.22 / Avg: 120.26 / Max: 126.65 Min: 85.74 / Avg: 126.1 / Max: 134.53 Min: 97.89 / Avg: 192.93 / Max: 216.11 Min: 41.92 / Avg: 196.37 / Max: 200.73 Min: 171.99 / Avg: 232.4 / Max: 318.93 Min: 56.74 / Avg: 235.38 / Max: 240.55 Min: 78.38 / Avg: 248.23 / Max: 290.1 Min: 60.77 / Avg: 261.84 / Max: 267.2 Min: 137.33 / Avg: 281.43 / Max: 342.59 Min: 124.02 / Avg: 284.83 / Max: 338.05
ClickHouse ClickHouse is an open-source, high performance OLAP data management system. This test profile uses ClickHouse's standard benchmark recommendations per https://clickhouse.com/docs/en/operations/performance-test/ / https://github.com/ClickHouse/ClickBench/tree/main/clickhouse with the 100 million rows web analytics dataset. The reported value is the query processing time using the geometric mean of all separate queries performed as an aggregate. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, Third Run EPYC 9684X EPYC 9654 EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon Platinum 8380 EPYC 8534P Xeon 6780E 2P EPYC 8534PN Xeon 6766E 2P Xeon Platinum 8380 2P 130 260 390 520 650 SE +/- 4.34, N = 3 SE +/- 4.03, N = 3 SE +/- 2.03, N = 3 SE +/- 10.00, N = 3 SE +/- 2.55, N = 3 SE +/- 3.08, N = 3 SE +/- 3.45, N = 3 SE +/- 3.60, N = 3 SE +/- 3.54, N = 12 SE +/- 6.03, N = 3 SE +/- 3.32, N = 3 SE +/- 3.62, N = 6 SE +/- 5.39, N = 5 SE +/- 6.21, N = 5 SE +/- 1.22, N = 3 SE +/- 2.14, N = 3 SE +/- 4.79, N = 3 SE +/- 1.77, N = 3 SE +/- 3.52, N = 3 SE +/- 5.23, N = 3 SE +/- 1.59, N = 3 SE +/- 1.27, N = 3 595.19 559.55 526.36 525.23 501.63 501.07 481.66 476.91 471.79 470.02 458.96 458.23 438.60 427.95 400.52 379.92 342.99 338.84 337.75 335.46 320.87 295.06 Queries Per Minute, Geo Mean Per Watt
OpenBenchmarking.org Queries Per Minute, Geo Mean Per Watt, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, Third Run Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 9754 EPYC 8534P EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 EPYC 9654 2P EPYC 9754 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Max 9480 2P 1.1063 2.2126 3.3189 4.4252 5.5315 4.917 4.737 4.366 4.182 4.085 4.053 3.709 2.599 2.341 2.049 2.038 1.974 1.967 1.966 1.870 1.857 1.676 1.289 1.265 1.068 1.043 1.012 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ClickHouse 22.12.3.5 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 EPYC 9654 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 15.83 / Avg: 76.83 / Max: 137.9 Min: 43.57 / Avg: 77.27 / Max: 231.46 Min: 18.06 / Avg: 82.96 / Max: 166.09 Min: 43.42 / Avg: 84.55 / Max: 263.64 Min: 20.77 / Avg: 109.76 / Max: 265.96 Min: 28.3 / Avg: 138.04 / Max: 305.03 Min: 62.49 / Avg: 157.43 / Max: 363.95 Min: 29.83 / Avg: 160.48 / Max: 343.21 Min: 107.74 / Avg: 164.81 / Max: 404.79 Min: 52.36 / Avg: 174.43 / Max: 288.3 Min: 26.54 / Avg: 176.29 / Max: 377.99 Min: 52.84 / Avg: 205.71 / Max: 375.59 Min: 40.81 / Avg: 230.4 / Max: 456.57 Min: 70.28 / Avg: 238.9 / Max: 367.38 Min: 77.89 / Avg: 241.61 / Max: 372.85 Min: 54.72 / Avg: 280.91 / Max: 531.55 Min: 101.69 / Avg: 282.88 / Max: 572.19 Min: 64.74 / Avg: 314 / Max: 643.44 Min: 77.74 / Avg: 340.3 / Max: 642.4 Min: 157.92 / Avg: 396.39 / Max: 673.71 Min: 199.33 / Avg: 466.07 / Max: 698.95 Min: 199.78 / Avg: 469.37 / Max: 713.86
OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, Second Run EPYC 9684X EPYC 9654 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Platinum 8592+ EPYC 9754 Xeon Platinum 8592+ 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon Platinum 8380 Xeon 6780E 2P EPYC 8534PN EPYC 8534P Xeon 6766E 2P Xeon Platinum 8380 2P 130 260 390 520 650 SE +/- 0.42, N = 3 SE +/- 3.14, N = 3 SE +/- 8.37, N = 3 SE +/- 6.78, N = 3 SE +/- 3.26, N = 3 SE +/- 0.19, N = 3 SE +/- 2.93, N = 3 SE +/- 4.95, N = 3 SE +/- 3.17, N = 12 SE +/- 1.94, N = 3 SE +/- 2.86, N = 6 SE +/- 9.26, N = 3 SE +/- 3.10, N = 5 SE +/- 4.85, N = 5 SE +/- 2.12, N = 3 SE +/- 3.34, N = 3 SE +/- 0.73, N = 3 SE +/- 3.25, N = 3 SE +/- 5.72, N = 3 SE +/- 2.99, N = 3 SE +/- 1.53, N = 3 SE +/- 3.61, N = 3 597.73 562.50 529.80 528.21 507.56 499.18 486.05 474.68 467.96 465.06 455.79 450.12 443.58 420.35 403.96 381.05 339.67 339.50 337.44 337.32 320.17 295.43
OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, First Run / Cold Cache EPYC 9684X EPYC 9654 EPYC 9684X 2P EPYC 9654 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Platinum 8592+ EPYC 9754 Xeon Platinum 8592+ 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon Platinum 8380 EPYC 8534P EPYC 8534PN Xeon 6766E 2P Xeon Platinum 8380 2P 120 240 360 480 600 SE +/- 7.88, N = 3 SE +/- 5.33, N = 3 SE +/- 4.34, N = 3 SE +/- 6.05, N = 3 SE +/- 1.35, N = 3 SE +/- 4.37, N = 3 SE +/- 3.41, N = 3 SE +/- 5.88, N = 3 SE +/- 3.85, N = 12 SE +/- 2.85, N = 3 SE +/- 4.18, N = 6 SE +/- 2.78, N = 3 SE +/- 4.12, N = 5 SE +/- 4.49, N = 5 SE +/- 1.12, N = 3 SE +/- 2.21, N = 3 SE +/- 3.71, N = 3 SE +/- 1.56, N = 3 SE +/- 1.57, N = 3 SE +/- 1.88, N = 3 SE +/- 1.51, N = 3 SE +/- 1.56, N = 3 566.31 536.82 513.92 504.84 491.89 485.63 465.50 461.62 460.66 446.40 437.20 433.79 423.30 411.26 388.29 365.62 333.06 326.65 323.59 322.69 311.63 285.17
Apache IoTDB Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400 EPYC 9684X EPYC 9654 EPYC 9754 EPYC 9654 2P EPYC 9684X 2P EPYC 8534PN EPYC 8534P EPYC 9754 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon 6780E Xeon Max 9480 Xeon 6766E Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Platinum 8380 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon 6780E 2P Xeon 6766E 2P 80 160 240 320 400 SE +/- 1.71, N = 3 SE +/- 1.96, N = 3 SE +/- 1.41, N = 3 SE +/- 3.49, N = 3 SE +/- 1.87, N = 3 SE +/- 0.82, N = 3 SE +/- 0.38, N = 3 SE +/- 2.16, N = 15 SE +/- 3.41, N = 3 SE +/- 3.40, N = 12 SE +/- 2.20, N = 3 SE +/- 4.45, N = 3 SE +/- 5.07, N = 3 SE +/- 1.12, N = 3 SE +/- 5.15, N = 3 SE +/- 3.79, N = 8 SE +/- 3.74, N = 12 SE +/- 4.06, N = 3 SE +/- 2.59, N = 12 SE +/- 3.47, N = 3 SE +/- 5.20, N = 3 SE +/- 3.93, N = 12 209.68 211.85 227.40 240.43 246.08 251.05 251.47 253.13 280.26 286.78 294.33 296.36 309.03 312.32 314.44 317.36 325.97 326.53 328.41 349.42 377.58 380.33 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Apache IoTDB 1.2 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8380 Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 120 240 360 480 600 Min: 17.44 / Avg: 68.28 / Max: 132.65 Min: 15.92 / Avg: 72.92 / Max: 146.34 Min: 43.59 / Avg: 76.38 / Max: 156.34 Min: 43.44 / Avg: 82.74 / Max: 200.76 Min: 21.99 / Avg: 100.77 / Max: 212.81 Min: 30.38 / Avg: 122.12 / Max: 263.01 Min: 35.8 / Avg: 141.12 / Max: 306.16 Min: 33.32 / Avg: 159.48 / Max: 348.24 Min: 51.86 / Avg: 163.52 / Max: 286.64 Min: 61.74 / Avg: 182.78 / Max: 332.16 Min: 95.98 / Avg: 192.34 / Max: 351.09 Min: 58.74 / Avg: 204.83 / Max: 345.97 Min: 43.37 / Avg: 215.97 / Max: 369.95 Min: 75.58 / Avg: 242.09 / Max: 358.23 Min: 74.42 / Avg: 245.7 / Max: 359.89 Min: 52.66 / Avg: 254.51 / Max: 448.61 Min: 68.46 / Avg: 291.76 / Max: 496.27 Min: 105.33 / Avg: 293.48 / Max: 476.12 Min: 76.05 / Avg: 329.73 / Max: 605.38 Min: 174.02 / Avg: 422.43 / Max: 631.3 Min: 207 / Avg: 485.64 / Max: 662.26 Min: 204.98 / Avg: 492.49 / Max: 690.42
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400 EPYC 9684X EPYC 9654 EPYC 9754 EPYC 9654 2P EPYC 9684X 2P EPYC 8534P EPYC 8534PN EPYC 9754 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon 6780E Xeon Max 9480 Xeon Max 9468 Xeon 6766E Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon 6780E 2P Xeon 6766E 2P 30M 60M 90M 120M 150M SE +/- 371659.60, N = 3 SE +/- 255887.16, N = 3 SE +/- 68559.13, N = 3 SE +/- 1152766.29, N = 3 SE +/- 353097.82, N = 3 SE +/- 13050.78, N = 3 SE +/- 412757.79, N = 3 SE +/- 666882.79, N = 15 SE +/- 746822.47, N = 3 SE +/- 761276.42, N = 12 SE +/- 308756.34, N = 3 SE +/- 441557.52, N = 3 SE +/- 1037167.36, N = 3 SE +/- 662950.65, N = 3 SE +/- 111181.55, N = 3 SE +/- 692701.79, N = 8 SE +/- 88241.50, N = 3 SE +/- 574699.97, N = 12 SE +/- 565297.90, N = 12 SE +/- 661234.22, N = 3 SE +/- 796440.21, N = 3 SE +/- 531367.14, N = 12 126161028 125562060 116685326 108821580 106778379 105432862 105138551 103630483 92709751 89231850 87797206 87190397 83475645 82455664 81900477 80169620 78823112 78057728 77501248 74653007 68320800 67599793
OpenRadioss OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: Chrysler Neon 1M EPYC 9684X 2P EPYC 9684X EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9754 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 9654 Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9754 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 70 140 210 280 350 SE +/- 0.46, N = 3 SE +/- 0.49, N = 3 SE +/- 0.71, N = 12 SE +/- 0.14, N = 3 SE +/- 1.00, N = 10 SE +/- 0.11, N = 3 SE +/- 0.14, N = 3 SE +/- 0.13, N = 3 SE +/- 0.20, N = 3 SE +/- 1.51, N = 8 SE +/- 0.78, N = 3 SE +/- 0.15, N = 3 SE +/- 0.18, N = 3 SE +/- 0.08, N = 3 SE +/- 1.11, N = 3 SE +/- 0.47, N = 3 SE +/- 0.50, N = 3 SE +/- 0.13, N = 3 49.55 77.32 77.64 86.93 93.99 115.89 117.91 125.15 157.56 176.14 185.03 207.20 212.99 228.74 231.96 297.25 300.69 343.16 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenRadioss 2023.09.15 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 Xeon Platinum 8380 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 17.52 / Avg: 128.33 / Max: 140.79 Min: 18.8 / Avg: 143.22 / Max: 157.87 Min: 21.25 / Avg: 249.04 / Max: 321.21 Min: 52.15 / Avg: 250.92 / Max: 289.23 Min: 28.56 / Avg: 270.92 / Max: 332.67 Min: 33.17 / Avg: 309.02 / Max: 409.54 Min: 33.32 / Avg: 313.77 / Max: 375.72 Min: 59.23 / Avg: 321.19 / Max: 377.05 Min: 72.8 / Avg: 328.51 / Max: 377.04 Min: 83.35 / Avg: 329.11 / Max: 376.3 Min: 42.16 / Avg: 450.48 / Max: 658.58 Min: 105.1 / Avg: 469.12 / Max: 579.49 Min: 59.6 / Avg: 495.69 / Max: 689.03 Min: 68.17 / Avg: 526.7 / Max: 819.28 Min: 82.45 / Avg: 561.7 / Max: 752.76 Min: 168.54 / Avg: 600.33 / Max: 753.81 Min: 213.51 / Avg: 619.95 / Max: 751.65 Min: 197.55 / Avg: 624.98 / Max: 753.94
NAMD NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ns/day, More Is Better NAMD 3.0b6 Input: STMV with 1,066,628 Atoms EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P EPYC 9684X EPYC 9754 EPYC 9654 Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8380 2P EPYC 8534P Xeon Max 9480 EPYC 8534PN Xeon Max 9468 Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon 6766E 2P Xeon 6780E 2P Xeon 6780E Xeon 6766E 2 4 6 8 10 SE +/- 0.01739, N = 3 SE +/- 0.00332, N = 3 SE +/- 0.02927, N = 3 SE +/- 0.00206, N = 3 SE +/- 0.00847, N = 3 SE +/- 0.00537, N = 3 SE +/- 0.08185, N = 13 SE +/- 0.05644, N = 12 SE +/- 0.05463, N = 15 SE +/- 0.00911, N = 3 SE +/- 0.00440, N = 3 SE +/- 0.05862, N = 15 SE +/- 0.00540, N = 3 SE +/- 0.05087, N = 15 SE +/- 0.02648, N = 15 SE +/- 0.00653, N = 3 SE +/- 0.04403, N = 15 SE +/- 0.01396, N = 15 SE +/- 0.04439, N = 8 SE +/- 0.03216, N = 9 SE +/- 0.00540, N = 3 SE +/- 0.00423, N = 3 6.38343 5.83606 5.27839 3.41085 2.88869 2.87877 2.25422 1.88679 1.76778 1.61232 1.51901 1.50214 1.46163 1.45223 1.43787 1.11056 1.10900 1.05274 0.70757 0.57694 0.56717 0.52357 ns/day Per Watt
OpenBenchmarking.org ns/day Per Watt, More Is Better NAMD 3.0b6 Input: STMV with 1,066,628 Atoms EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P EPYC 9754 EPYC 8534PN EPYC 8534P EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Max 9480 Xeon Max 9468 Xeon 6780E Xeon 6766E Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon 6766E 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E 2P Xeon Platinum 8490H 2P 0.0052 0.0104 0.0156 0.0208 0.026 0.023 0.019 0.018 0.014 0.013 0.012 0.012 0.012 0.006 0.005 0.005 0.005 0.004 0.004 0.004 0.004 0.004 0.003 0.003 0.003 0.002 0.002 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better NAMD 3.0b6 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon 6766E 2P EPYC 9754 2P Xeon Platinum 8592+ Xeon 6780E 2P Xeon Platinum 8490H EPYC 9684X Xeon Max 9468 EPYC 9654 2P Xeon Max 9480 EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 15.81 / Avg: 114.62 / Max: 140.3 Min: 43.64 / Avg: 119.15 / Max: 229.24 Min: 18.33 / Avg: 130.18 / Max: 164.25 Min: 43.54 / Avg: 140.31 / Max: 283.29 Min: 21.48 / Avg: 206.2 / Max: 309.47 Min: 51.87 / Avg: 233.31 / Max: 276.8 Min: 28.55 / Avg: 235.77 / Max: 330 Min: 104.7 / Avg: 241.31 / Max: 387.86 Min: 42.36 / Avg: 249.94 / Max: 651.86 Min: 33.09 / Avg: 260.62 / Max: 349.3 Min: 62.83 / Avg: 273.53 / Max: 463.21 Min: 55.23 / Avg: 273.91 / Max: 344.33 Min: 32.28 / Avg: 274.76 / Max: 406.72 Min: 71.79 / Avg: 281.33 / Max: 364.12 Min: 57.51 / Avg: 289.28 / Max: 672.46 Min: 72.24 / Avg: 295.54 / Max: 371.41 Min: 65.49 / Avg: 343.59 / Max: 811.63 Min: 103.57 / Avg: 411.33 / Max: 527.15 Min: 73.13 / Avg: 510.74 / Max: 686.96 Min: 163.14 / Avg: 552.42 / Max: 653.46 Min: 192.28 / Avg: 560.13 / Max: 714.79 Min: 194.25 / Avg: 565.07 / Max: 686.05
Timed Gem5 Compilation This test times how long it takes to compile Gem5. Gem5 is a simulator for computer system architecture research. Gem5 is widely used for computer architecture research within the industry, academia, and more. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Timed Gem5 Compilation 23.0.1 Time To Compile Xeon Platinum 8592+ 2P EPYC 9684X 2P EPYC 9654 2P EPYC 9684X Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ EPYC 9754 2P Xeon Platinum 8490H EPYC 9754 Xeon Max 9480 Xeon Platinum 8380 2P Xeon Max 9468 Xeon 6780E 2P Xeon 6780E EPYC 8534P EPYC 8534PN Xeon 6766E 2P Xeon 6766E Xeon Platinum 8380 60 120 180 240 300 SE +/- 0.89, N = 3 SE +/- 1.16, N = 3 SE +/- 0.72, N = 3 SE +/- 2.15, N = 3 SE +/- 0.58, N = 3 SE +/- 1.01, N = 3 SE +/- 1.72, N = 3 SE +/- 2.24, N = 4 SE +/- 2.17, N = 3 SE +/- 1.73, N = 3 SE +/- 0.92, N = 3 SE +/- 1.88, N = 3 SE +/- 2.11, N = 6 SE +/- 1.04, N = 3 SE +/- 2.59, N = 4 SE +/- 2.61, N = 3 SE +/- 1.47, N = 3 SE +/- 1.99, N = 9 SE +/- 0.97, N = 3 SE +/- 0.78, N = 3 SE +/- 3.60, N = 3 SE +/- 3.88, N = 3 157.36 161.94 164.90 176.88 178.28 181.63 184.30 185.30 187.51 192.98 210.93 212.55 215.50 219.67 220.57 233.21 238.43 242.93 251.61 252.97 267.95 275.71 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Timed Gem5 Compilation 23.0.1 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 Xeon 6766E 2P EPYC 9654 Xeon 6780E 2P Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 16.66 / Avg: 87.52 / Max: 141.58 Min: 43.55 / Avg: 90.66 / Max: 234.37 Min: 17.32 / Avg: 96.65 / Max: 166.6 Min: 43.31 / Avg: 103.5 / Max: 281.26 Min: 21.18 / Avg: 144.58 / Max: 331.44 Min: 104.15 / Avg: 164.39 / Max: 501.65 Min: 29.61 / Avg: 170.81 / Max: 332.22 Min: 55.47 / Avg: 178.9 / Max: 585.97 Min: 50.89 / Avg: 194.29 / Max: 291.11 Min: 29.1 / Avg: 201.66 / Max: 413.09 Min: 33.1 / Avg: 215.42 / Max: 377.33 Min: 55.65 / Avg: 228.11 / Max: 376.67 Min: 67.72 / Avg: 254.21 / Max: 375.35 Min: 40.79 / Avg: 257.31 / Max: 376.35 Min: 40.43 / Avg: 270.24 / Max: 677.31 Min: 55.44 / Avg: 320.45 / Max: 676.84 Min: 97.98 / Avg: 327.07 / Max: 580.5 Min: 65.57 / Avg: 360.36 / Max: 822.7 Min: 43.96 / Avg: 377.81 / Max: 753.32 Min: 160.34 / Avg: 415.56 / Max: 738.65 Min: 191.44 / Avg: 472.13 / Max: 747.43 Min: 177.84 / Avg: 479.92 / Max: 746.22
LAMMPS Molecular Dynamics Simulator LAMMPS is a classical molecular dynamics code, and an acronym for Large-scale Atomic/Molecular Massively Parallel Simulator. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ns/day, More Is Better LAMMPS Molecular Dynamics Simulator 23Jun2022 Model: 20k Atoms EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P EPYC 9754 EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 14 28 42 56 70 SE +/- 0.12, N = 3 SE +/- 0.13, N = 3 SE +/- 0.13, N = 3 SE +/- 0.04, N = 3 SE +/- 0.05, N = 3 SE +/- 0.01, N = 3 SE +/- 0.19, N = 3 SE +/- 0.30, N = 3 SE +/- 0.07, N = 3 SE +/- 0.12, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 62.00 58.84 57.10 56.40 51.82 50.52 47.99 46.89 46.71 42.51 37.46 35.70 34.89 32.16 28.28 27.53 25.33 21.38 1. (CXX) g++ options: -O3 -lm -ldl ns/day Per Watt
OpenBenchmarking.org ns/day Per Watt, More Is Better LAMMPS Molecular Dynamics Simulator 23Jun2022 Model: 20k Atoms EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9480 EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Max 9468 Xeon Platinum 8380 EPYC 9684X 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Platinum 8380 2P 0.0432 0.0864 0.1296 0.1728 0.216 0.192 0.180 0.151 0.133 0.117 0.110 0.102 0.094 0.093 0.090 0.084 0.082 0.080 0.079 0.076 0.074 0.068 0.067 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better LAMMPS Molecular Dynamics Simulator 23Jun2022 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 15.43 / Avg: 132.09 / Max: 137.27 Min: 15.99 / Avg: 153.16 / Max: 156.26 Min: 52.92 / Avg: 266.98 / Max: 289.96 Min: 19.76 / Avg: 318.23 / Max: 330.98 Min: 29.12 / Avg: 318.64 / Max: 329.98 Min: 22.64 / Avg: 341.9 / Max: 376.37 Min: 61.2 / Avg: 343.29 / Max: 376.01 Min: 77.1 / Avg: 344.05 / Max: 376.79 Min: 75.25 / Avg: 345.23 / Max: 374.68 Min: 31.69 / Avg: 399.48 / Max: 409.61 Min: 97.85 / Avg: 530.19 / Max: 581.83 Min: 40.67 / Avg: 624.63 / Max: 672.65 Min: 55.66 / Avg: 633.05 / Max: 669.84 Min: 65.48 / Avg: 673.08 / Max: 752.97 Min: 165.47 / Avg: 680.62 / Max: 754.38 Min: 192.84 / Avg: 682.64 / Max: 752.01 Min: 186.6 / Avg: 685.33 / Max: 753.98 Min: 64.31 / Avg: 785.79 / Max: 819.57
Timed Node.js Compilation This test profile times how long it takes to build/compile Node.js itself from source. Node.js is a JavaScript run-time built from the Chrome V8 JavaScript engine while itself is written in C/C++. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Timed Node.js Compilation 21.7.2 Time To Compile EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9754 2P EPYC 9684X EPYC 9654 Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon Max 9468 2P EPYC 9754 Xeon 6766E 2P Xeon Platinum 8592+ Xeon 6780E Xeon 6766E Xeon Max 9480 Xeon Platinum 8490H Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon Platinum 8380 90 180 270 360 450 SE +/- 1.19, N = 3 SE +/- 0.18, N = 3 SE +/- 0.12, N = 3 SE +/- 1.51, N = 3 SE +/- 0.50, N = 3 SE +/- 0.59, N = 3 SE +/- 0.52, N = 3 SE +/- 0.48, N = 3 SE +/- 0.60, N = 3 SE +/- 0.84, N = 3 SE +/- 0.72, N = 3 SE +/- 1.20, N = 3 SE +/- 0.52, N = 3 SE +/- 0.51, N = 3 SE +/- 0.15, N = 3 SE +/- 0.74, N = 3 SE +/- 0.71, N = 3 SE +/- 0.53, N = 3 SE +/- 0.17, N = 3 SE +/- 0.19, N = 3 SE +/- 0.68, N = 3 SE +/- 0.22, N = 3 114.91 118.90 145.41 147.74 147.84 164.02 171.17 172.99 173.09 175.63 186.10 187.65 214.43 228.57 246.85 252.65 254.00 270.38 277.53 288.12 303.42 423.63 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Timed Node.js Compilation 21.7.2 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 Xeon 6766E 2P Xeon 6780E 2P EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 16.32 / Avg: 118.09 / Max: 147.17 Min: 18.11 / Avg: 134.67 / Max: 172.23 Min: 43.59 / Avg: 148.97 / Max: 246.66 Min: 43.35 / Avg: 170.44 / Max: 300.92 Min: 21.84 / Avg: 210.13 / Max: 338.39 Min: 29.13 / Avg: 242.47 / Max: 351.76 Min: 50.7 / Avg: 248.06 / Max: 290.96 Min: 104.39 / Avg: 256.25 / Max: 503.96 Min: 106.88 / Avg: 287.54 / Max: 584 Min: 30.95 / Avg: 289.15 / Max: 411.31 Min: 33.43 / Avg: 295.74 / Max: 377.74 Min: 56.59 / Avg: 305.63 / Max: 376.82 Min: 84.71 / Avg: 317.34 / Max: 376.11 Min: 74.34 / Avg: 318.88 / Max: 376.31 Min: 41.45 / Avg: 361.3 / Max: 676 Min: 55.68 / Avg: 430.05 / Max: 696.5 Min: 96.84 / Avg: 455.83 / Max: 582.18 Min: 61.4 / Avg: 490.25 / Max: 822.63 Min: 82.39 / Avg: 529.15 / Max: 756.15 Min: 157.7 / Avg: 555.35 / Max: 756.3 Min: 190.08 / Avg: 584.7 / Max: 754.37 Min: 200.82 / Avg: 601.52 / Max: 753.5
OpenSSL OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org byte/s, More Is Better OpenSSL 3.3 Algorithm: SHA256 EPYC 9754 2P Xeon 6780E 2P EPYC 9684X 2P Xeon 6766E 2P EPYC 9654 2P EPYC 9754 Xeon 6780E EPYC 9684X Xeon Platinum 8592+ 2P Xeon 6766E EPYC 9654 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 70000M 140000M 210000M 280000M 350000M SE +/- 80639802.67, N = 3 SE +/- 228931567.43, N = 3 SE +/- 160246527.97, N = 3 SE +/- 97765007.32, N = 3 SE +/- 210407867.05, N = 3 SE +/- 34182873.86, N = 3 SE +/- 242095363.59, N = 3 SE +/- 146384349.91, N = 3 SE +/- 29142783.86, N = 3 SE +/- 34216643.88, N = 3 SE +/- 38601729.93, N = 3 SE +/- 66583540.43, N = 3 SE +/- 70647143.49, N = 3 SE +/- 126029257.64, N = 3 SE +/- 10642001.84, N = 3 SE +/- 64797308.82, N = 3 SE +/- 23323764.84, N = 3 SE +/- 159515825.53, N = 3 SE +/- 919807437.76, N = 12 SE +/- 33334799.97, N = 3 SE +/- 27466135.54, N = 3 SE +/- 8534071.27, N = 3 320258686237 287035041140 276267766403 256330349777 251645629887 161269162280 144163027280 138409588983 131976050120 129975994733 126475137690 119106802107 103382860113 88635753413 74584721007 67806730387 66573957920 61806994523 54530700576 51840436763 44439185187 27858504490 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl byte/s Per Watt
OpenBenchmarking.org byte/s Per Watt, More Is Better OpenSSL 3.3 Algorithm: SHA256 Xeon 6766E Xeon 6766E 2P Xeon 6780E EPYC 8534PN Xeon 6780E 2P EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 9654 EPYC 9654 2P EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 130M 260M 390M 520M 650M 611047359.48 578313018.50 516781889.88 511227929.81 507731880.45 481947170.76 471857277.56 469088919.27 386445411.79 378210579.19 345260515.93 344443539.20 195213927.96 193109819.08 189585676.93 179424218.79 167457577.18 160789835.78 157500377.52 149911179.45 105324899.05 103421352.66 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H EPYC 9654 EPYC 9754 Xeon Platinum 8592+ EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.53 / Avg: 132.64 / Max: 140.44 Min: 18.27 / Avg: 159 / Max: 170.75 Min: 43.63 / Avg: 212.71 / Max: 270.71 Min: 52.69 / Avg: 264.5 / Max: 291.33 Min: 43.49 / Avg: 278.96 / Max: 357.32 Min: 75.75 / Avg: 282.15 / Max: 323.37 Min: 82.11 / Avg: 309.57 / Max: 363.52 Min: 57.24 / Avg: 326.01 / Max: 380.05 Min: 29.19 / Avg: 327.28 / Max: 361.5 Min: 22.61 / Avg: 334.62 / Max: 353.13 Min: 32.34 / Avg: 341.03 / Max: 377.49 Min: 34.63 / Avg: 400.88 / Max: 411.05 Min: 105.85 / Avg: 443.24 / Max: 540.57 Min: 100.31 / Avg: 527.27 / Max: 585.66 Min: 108.39 / Avg: 565.33 / Max: 713.33 Min: 182.49 / Avg: 591.26 / Max: 697.86 Min: 186.31 / Avg: 642.97 / Max: 734.25 Min: 163.14 / Avg: 663.83 / Max: 754.31 Min: 59.61 / Avg: 665.36 / Max: 734.49 Min: 42.79 / Avg: 678.72 / Max: 712.84 Min: 71.86 / Avg: 683.42 / Max: 754.43 Min: 69.9 / Avg: 802.07 / Max: 822.12
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 512 - Model: GoogLeNet EPYC 9754 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9654 2P Xeon Max 9480 2P EPYC 9654 Xeon Max 9468 2P EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 EPYC 9684X Xeon Platinum 8380 2P Xeon Max 9480 Xeon Max 9468 Xeon 6780E 2P Xeon 6766E 2P EPYC 8534P Xeon 6780E EPYC 8534PN Xeon Platinum 8380 Xeon 6766E 130 260 390 520 650 SE +/- 5.63, N = 6 SE +/- 1.52, N = 3 SE +/- 2.62, N = 3 SE +/- 5.11, N = 3 SE +/- 2.46, N = 3 SE +/- 4.02, N = 10 SE +/- 1.14, N = 3 SE +/- 1.60, N = 3 SE +/- 3.68, N = 3 SE +/- 0.34, N = 3 SE +/- 12.03, N = 12 SE +/- 3.05, N = 12 SE +/- 0.25, N = 3 SE +/- 0.37, N = 3 SE +/- 0.41, N = 3 SE +/- 1.26, N = 3 SE +/- 1.19, N = 3 SE +/- 0.06, N = 3 SE +/- 0.19, N = 3 SE +/- 0.03, N = 3 SE +/- 0.12, N = 3 SE +/- 0.26, N = 3 588.43 577.76 576.08 525.84 477.33 470.95 467.91 464.78 462.78 461.23 450.61 425.25 411.05 398.40 383.57 378.79 354.81 313.50 298.16 295.29 290.16 278.05 images/sec Per Watt
OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 512 - Model: GoogLeNet EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6766E EPYC 9684X EPYC 9754 2P Xeon Max 9480 Xeon 6780E Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon 6766E 2P EPYC 9684X 2P Xeon 6780E 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.5085 1.017 1.5255 2.034 2.5425 2.260 2.098 1.824 1.646 1.511 1.378 1.298 1.293 1.245 1.180 1.140 1.137 1.110 0.974 0.960 0.871 0.813 0.781 0.747 0.734 0.710 0.696 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.16.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon Platinum 8380 Xeon 6780E EPYC 9654 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6766E 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 16.37 / Avg: 130.65 / Max: 139.18 Min: 18.68 / Avg: 149.41 / Max: 158.78 Min: 43.64 / Avg: 214.28 / Max: 238.23 Min: 19.9 / Avg: 247.07 / Max: 298.24 Min: 52.49 / Avg: 261.52 / Max: 277.08 Min: 43.63 / Avg: 261.65 / Max: 298.46 Min: 16.07 / Avg: 286.15 / Max: 318.45 Min: 35.37 / Avg: 306.26 / Max: 336.64 Min: 33.11 / Avg: 328.89 / Max: 367.06 Min: 62.15 / Avg: 334.67 / Max: 366.71 Min: 73.74 / Avg: 337.25 / Max: 358.42 Min: 83.34 / Avg: 337.62 / Max: 362.35 Min: 106.15 / Avg: 454.45 / Max: 503.93 Min: 41.61 / Avg: 472.75 / Max: 541.37 Min: 104.77 / Avg: 505.8 / Max: 538.69 Min: 109.13 / Avg: 515.79 / Max: 577.43 Min: 60.16 / Avg: 547.94 / Max: 597.52 Min: 81.09 / Avg: 593.47 / Max: 647.85 Min: 65.4 / Avg: 622.34 / Max: 677.73 Min: 163.66 / Avg: 661.11 / Max: 719.14 Min: 205.47 / Avg: 671.88 / Max: 722 Min: 204.77 / Avg: 672.55 / Max: 725.44
FFmpeg This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org FPS, More Is Better FFmpeg 7.0 Encoder: libx265 - Scenario: Platform EPYC 9654 EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9468 EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 8534PN Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 12 24 36 48 60 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 SE +/- 0.14, N = 3 SE +/- 0.15, N = 3 SE +/- 0.01, N = 3 SE +/- 0.16, N = 3 SE +/- 0.12, N = 3 SE +/- 0.08, N = 3 SE +/- 0.03, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.27, N = 3 SE +/- 0.09, N = 3 SE +/- 0.08, N = 3 SE +/- 0.05, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.03, N = 3 55.61 55.38 55.11 53.92 53.22 52.79 49.79 48.76 48.35 47.32 47.31 47.13 46.87 46.86 46.59 45.95 44.01 43.55 29.94 28.39 27.17 26.90 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma FPS Per Watt
OpenBenchmarking.org FPS Per Watt, More Is Better FFmpeg 7.0 Encoder: libx265 - Scenario: Platform EPYC 8534PN EPYC 8534P Xeon 6780E Xeon 6766E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P Xeon Max 9480 Xeon Max 9468 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.1575 0.315 0.4725 0.63 0.7875 0.700 0.675 0.516 0.498 0.498 0.452 0.399 0.329 0.303 0.267 0.230 0.224 0.222 0.212 0.210 0.203 0.192 0.184 0.168 0.133 0.106 0.104 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better FFmpeg 7.0 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6766E 2P EPYC 9654 Xeon 6780E 2P EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 90 180 270 360 450 Min: 43.59 / Avg: 54.55 / Max: 78.03 Min: 43.43 / Avg: 58 / Max: 86.79 Min: 15.45 / Avg: 66.97 / Max: 86.23 Min: 15.65 / Avg: 69.43 / Max: 88.73 Min: 20.65 / Avg: 94.97 / Max: 114.96 Min: 56.04 / Avg: 117.19 / Max: 147.2 Min: 27.62 / Avg: 123.14 / Max: 151.27 Min: 61.79 / Avg: 127.88 / Max: 169.4 Min: 28.08 / Avg: 138.95 / Max: 171.83 Min: 49.55 / Avg: 145.35 / Max: 194.36 Min: 27.23 / Avg: 163.84 / Max: 212.24 Min: 62.31 / Avg: 186.7 / Max: 222.86 Min: 39.9 / Avg: 210.51 / Max: 233.46 Min: 76.85 / Avg: 225.5 / Max: 260.64 Min: 81.77 / Avg: 228.19 / Max: 273.56 Min: 100.74 / Avg: 237.01 / Max: 295.05 Min: 57.48 / Avg: 260.25 / Max: 290.76 Min: 61.23 / Avg: 286.98 / Max: 323.54 Min: 68.77 / Avg: 316.56 / Max: 368.13 Min: 161.47 / Avg: 365.67 / Max: 399.89 Min: 212.12 / Avg: 439.49 / Max: 489.04 Min: 205.4 / Avg: 441.78 / Max: 509.36
Result
OpenBenchmarking.org FPS, More Is Better FFmpeg 7.0 Encoder: libx265 - Scenario: Video On Demand EPYC 9654 EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9468 EPYC 9754 EPYC 9754 2P EPYC 8534PN EPYC 8534P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 12 24 36 48 60 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.07, N = 3 SE +/- 0.03, N = 3 SE +/- 0.14, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.07, N = 3 SE +/- 0.10, N = 3 SE +/- 0.03, N = 3 SE +/- 0.00, N = 3 SE +/- 0.05, N = 3 SE +/- 0.02, N = 3 SE +/- 0.13, N = 3 SE +/- 0.43, N = 3 SE +/- 0.07, N = 3 SE +/- 0.02, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.04, N = 3 55.59 55.56 55.01 53.99 53.29 52.81 49.84 48.77 48.14 47.34 47.33 47.14 46.93 46.88 46.79 45.29 44.13 43.80 29.95 28.39 27.17 26.89 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma FPS Per Watt
OpenBenchmarking.org FPS Per Watt, More Is Better FFmpeg 7.0 Encoder: libx265 - Scenario: Video On Demand EPYC 8534PN EPYC 8534P Xeon 6780E EPYC 9754 Xeon 6766E EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P Xeon Max 9480 Xeon Max 9468 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.1577 0.3154 0.4731 0.6308 0.7885 0.701 0.675 0.516 0.499 0.498 0.452 0.400 0.329 0.303 0.267 0.229 0.224 0.222 0.211 0.210 0.203 0.191 0.184 0.168 0.133 0.107 0.102 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better FFmpeg 7.0 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6766E 2P EPYC 9654 Xeon 6780E 2P EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 90 180 270 360 450 Min: 43.59 / Avg: 54.51 / Max: 76.42 Min: 29 / Avg: 58.08 / Max: 84.77 Min: 16.1 / Avg: 66.97 / Max: 83.84 Min: 15.96 / Avg: 69.48 / Max: 89.87 Min: 19.72 / Avg: 94.92 / Max: 115.72 Min: 103.99 / Avg: 117.45 / Max: 147.92 Min: 28.98 / Avg: 123.07 / Max: 151.95 Min: 60.98 / Avg: 127.77 / Max: 168.5 Min: 33.72 / Avg: 139.07 / Max: 174.07 Min: 51.94 / Avg: 145.51 / Max: 200.88 Min: 27.6 / Avg: 164.12 / Max: 212.91 Min: 60.59 / Avg: 186.45 / Max: 226 Min: 40.07 / Avg: 210.54 / Max: 233.66 Min: 78.38 / Avg: 225.58 / Max: 260.03 Min: 85.24 / Avg: 228.49 / Max: 271.74 Min: 100.82 / Avg: 237.65 / Max: 291.56 Min: 56.36 / Avg: 260.19 / Max: 290.84 Min: 64.38 / Avg: 287.45 / Max: 321.56 Min: 70.84 / Avg: 317.18 / Max: 370.52 Min: 163.78 / Avg: 366.29 / Max: 401.74 Min: 214.26 / Avg: 439.08 / Max: 477.53 Min: 207.79 / Avg: 444.37 / Max: 498.92
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: fcn-resnet101-11 - Device: CPU - Executor: Standard Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9468 2P Xeon Max 9480 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 EPYC 9754 Xeon 6780E EPYC 9684X EPYC 9654 EPYC 9684X 2P Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P EPYC 9754 2P EPYC 8534P EPYC 9654 2P EPYC 8534PN 50 100 150 200 250 SE +/- 0.69, N = 15 SE +/- 0.44, N = 3 SE +/- 0.71, N = 3 SE +/- 1.55, N = 15 SE +/- 0.09, N = 3 SE +/- 5.82, N = 12 SE +/- 2.59, N = 15 SE +/- 3.36, N = 15 SE +/- 6.41, N = 15 SE +/- 2.32, N = 15 SE +/- 2.29, N = 15 SE +/- 0.21, N = 3 SE +/- 4.76, N = 12 SE +/- 4.84, N = 15 SE +/- 2.18, N = 15 SE +/- 0.13, N = 3 SE +/- 0.09, N = 3 SE +/- 0.06, N = 3 SE +/- 2.91, N = 3 SE +/- 6.87, N = 12 SE +/- 0.35, N = 3 SE +/- 7.19, N = 15 92.35 100.74 109.67 110.97 116.32 121.67 121.78 139.10 155.79 160.53 166.79 174.14 178.03 182.02 182.27 182.36 188.30 194.11 203.57 216.70 220.35 226.63 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E Xeon Platinum 8380 EPYC 9654 EPYC 9684X Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P 120 240 360 480 600 Min: 17.45 / Avg: 121.04 / Max: 134.63 Min: 19.31 / Avg: 140.51 / Max: 154.48 Min: 43.7 / Avg: 200.72 / Max: 216.16 Min: 21.08 / Avg: 240.98 / Max: 267.52 Min: 43.58 / Avg: 243.68 / Max: 263.63 Min: 53.6 / Avg: 252.85 / Max: 280.59 Min: 27.94 / Avg: 273.66 / Max: 300.92 Min: 33.76 / Avg: 321.88 / Max: 360.42 Min: 105.71 / Avg: 414.92 / Max: 451.7 Min: 41.83 / Avg: 453.96 / Max: 496.3 Min: 108.32 / Avg: 473.62 / Max: 514.95 Min: 100.08 / Avg: 498.61 / Max: 537.57 Min: 60.03 / Avg: 522.44 / Max: 571.64 Min: 64.5 / Avg: 611.91 / Max: 674.71
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: fcn-resnet101-11 - Device: CPU - Executor: Standard Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9468 2P Xeon Max 9480 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 EPYC 9754 Xeon 6780E EPYC 9684X EPYC 9654 EPYC 9684X 2P Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P EPYC 9754 2P EPYC 8534P EPYC 9654 2P EPYC 8534PN 3 6 9 12 15 SE +/- 0.07619, N = 15 SE +/- 0.04271, N = 3 SE +/- 0.05824, N = 3 SE +/- 0.11577, N = 15 SE +/- 0.00664, N = 3 SE +/- 0.34632, N = 12 SE +/- 0.16487, N = 15 SE +/- 0.15651, N = 15 SE +/- 0.22869, N = 15 SE +/- 0.08940, N = 15 SE +/- 0.08489, N = 15 SE +/- 0.00692, N = 3 SE +/- 0.16422, N = 12 SE +/- 0.15150, N = 15 SE +/- 0.06453, N = 15 SE +/- 0.00381, N = 3 SE +/- 0.00241, N = 3 SE +/- 0.00154, N = 3 SE +/- 0.07023, N = 3 SE +/- 0.14467, N = 12 SE +/- 0.00718, N = 3 SE +/- 0.11914, N = 15 10.83623 9.92684 9.11885 9.03367 8.59719 8.39962 8.26016 7.24187 6.54965 6.24740 6.01164 5.74234 5.66508 5.55018 5.49711 5.48343 5.31050 5.15166 4.91420 4.66491 4.53817 4.46533 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: fcn-resnet101-11 - Device: CPU - Executor: Standard Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9468 2P Xeon Max 9480 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 EPYC 9754 Xeon 6780E EPYC 9684X EPYC 9654 EPYC 9684X 2P Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P EPYC 9754 2P EPYC 8534P EPYC 9654 2P EPYC 8534PN 50 100 150 200 250 SE +/- 0.69, N = 15 SE +/- 0.44, N = 3 SE +/- 0.71, N = 3 SE +/- 1.55, N = 15 SE +/- 0.09, N = 3 SE +/- 5.82, N = 12 SE +/- 2.59, N = 15 SE +/- 3.36, N = 15 SE +/- 6.41, N = 15 SE +/- 2.32, N = 15 SE +/- 2.29, N = 15 SE +/- 0.21, N = 3 SE +/- 4.76, N = 12 SE +/- 4.84, N = 15 SE +/- 2.18, N = 15 SE +/- 0.13, N = 3 SE +/- 0.09, N = 3 SE +/- 0.06, N = 3 SE +/- 2.91, N = 3 SE +/- 6.87, N = 12 SE +/- 0.35, N = 3 SE +/- 7.19, N = 15 92.35 100.74 109.67 110.97 116.32 121.67 121.78 139.10 155.79 160.53 166.79 174.14 178.03 182.02 182.27 182.36 188.30 194.11 203.57 216.70 220.35 226.63 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
FFmpeg This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org FPS, More Is Better FFmpeg 7.0 Encoder: libx265 - Scenario: Upload EPYC 9654 EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 EPYC 9754 Xeon Max 9468 Xeon Max 9480 2P EPYC 9754 2P EPYC 8534PN EPYC 8534P Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 7 14 21 28 35 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.10, N = 3 SE +/- 0.07, N = 3 SE +/- 0.03, N = 3 SE +/- 0.00, N = 3 SE +/- 0.07, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 SE +/- 0.12, N = 3 SE +/- 0.07, N = 3 SE +/- 0.06, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 27.66 27.61 27.43 26.93 26.80 26.41 24.97 24.56 23.77 23.36 23.33 23.30 23.28 23.27 23.25 22.34 21.96 21.73 15.47 14.87 14.01 13.90 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma FPS Per Watt
OpenBenchmarking.org FPS Per Watt, More Is Better FFmpeg 7.0 Encoder: libx265 - Scenario: Upload EPYC 8534PN EPYC 8534P Xeon 6780E Xeon 6766E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P Xeon Max 9480 Xeon Max 9468 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.0826 0.1652 0.2478 0.3304 0.413 0.367 0.354 0.283 0.270 0.258 0.237 0.210 0.179 0.167 0.142 0.121 0.121 0.113 0.110 0.108 0.104 0.098 0.097 0.095 0.069 0.054 0.052 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better FFmpeg 7.0 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6766E 2P EPYC 9654 Xeon 6780E 2P Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 90 180 270 360 450 Min: 21.91 / Avg: 51.98 / Max: 83.48 Min: 26.8 / Avg: 54.72 / Max: 93.15 Min: 16.42 / Avg: 63.46 / Max: 99.36 Min: 17.2 / Avg: 65.65 / Max: 103.63 Min: 21.11 / Avg: 90.52 / Max: 130.74 Min: 103.98 / Avg: 114.66 / Max: 149.44 Min: 28.66 / Avg: 116.46 / Max: 162.4 Min: 105.83 / Avg: 122.91 / Max: 188.11 Min: 54.29 / Avg: 131.33 / Max: 221.85 Min: 32.37 / Avg: 131.52 / Max: 197.98 Min: 33.49 / Avg: 150.78 / Max: 267.25 Min: 59.92 / Avg: 175.77 / Max: 260.15 Min: 40.21 / Avg: 205.67 / Max: 248.29 Min: 77.87 / Avg: 215.12 / Max: 282.66 Min: 73.34 / Avg: 216.54 / Max: 288.18 Min: 101.47 / Avg: 224.76 / Max: 314.49 Min: 57.5 / Avg: 253.71 / Max: 316.9 Min: 60.82 / Avg: 280.31 / Max: 349.69 Min: 74.8 / Avg: 282.67 / Max: 421.23 Min: 169.76 / Avg: 355.39 / Max: 439.92 Min: 200.89 / Avg: 428.04 / Max: 517.82 Min: 200.59 / Avg: 429.86 / Max: 523.1
PostgreSQL This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Write - Average Latency Xeon 6780E EPYC 9654 EPYC 9684X Xeon 6766E Xeon Platinum 8592+ Xeon Platinum 8380 2P EPYC 9754 Xeon Max 9480 Xeon Platinum 8490H Xeon Max 9468 Xeon Platinum 8490H 2P EPYC 8534P Xeon Platinum 8380 Xeon Max 9480 2P EPYC 9754 2P Xeon Max 9468 2P EPYC 8534PN EPYC 9654 2P EPYC 9684X 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8592+ 2P 7 14 21 28 35 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.12, N = 3 SE +/- 0.05, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.12, N = 12 SE +/- 0.12, N = 3 SE +/- 0.02, N = 3 SE +/- 0.06, N = 3 SE +/- 0.10, N = 3 SE +/- 0.25, N = 12 SE +/- 0.32, N = 12 SE +/- 0.15, N = 3 SE +/- 0.14, N = 3 SE +/- 0.10, N = 3 11.54 12.50 12.58 12.63 13.26 13.69 13.84 13.85 14.04 14.29 14.86 16.32 16.34 16.66 16.88 16.96 18.05 18.06 18.16 26.29 28.81 30.20 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better PostgreSQL 16 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 Xeon 6766E 2P EPYC 9654 Xeon 6780E 2P Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8592+ 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 110 220 330 440 550 Min: 15.4 / Avg: 110.68 / Max: 134.33 Min: 16.91 / Avg: 135.16 / Max: 158.27 Min: 43.52 / Avg: 150.61 / Max: 191.14 Min: 43.39 / Avg: 172.92 / Max: 212.03 Min: 21.27 / Avg: 207.74 / Max: 251.71 Min: 54.49 / Avg: 239.87 / Max: 290.09 Min: 104.31 / Avg: 242.14 / Max: 314.19 Min: 19.41 / Avg: 254.31 / Max: 305.58 Min: 107.64 / Avg: 262.99 / Max: 345.55 Min: 31.89 / Avg: 275.46 / Max: 347.94 Min: 32.63 / Avg: 299.03 / Max: 361.05 Min: 54.62 / Avg: 299.36 / Max: 354.29 Min: 69.55 / Avg: 311.42 / Max: 350.84 Min: 86.13 / Avg: 315.59 / Max: 351.55 Min: 71.52 / Avg: 361.27 / Max: 515.84 Min: 37.64 / Avg: 361.97 / Max: 439.12 Min: 48.63 / Avg: 431.32 / Max: 519.43 Min: 103.15 / Avg: 454.63 / Max: 533.63 Min: 42.76 / Avg: 483.2 / Max: 581.92 Min: 162.26 / Avg: 519.08 / Max: 620.62 Min: 201.21 / Avg: 553.8 / Max: 622.91 Min: 197.27 / Avg: 561.38 / Max: 632.44
OpenBenchmarking.org TPS, More Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Write Xeon 6780E EPYC 9654 EPYC 9684X Xeon 6766E Xeon Platinum 8592+ Xeon Platinum 8380 2P EPYC 9754 Xeon Max 9480 Xeon Platinum 8490H Xeon Max 9468 Xeon Platinum 8490H 2P EPYC 8534P Xeon Platinum 8380 Xeon Max 9480 2P EPYC 9754 2P Xeon Max 9468 2P EPYC 9654 2P EPYC 8534PN EPYC 9684X 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8592+ 2P 20K 40K 60K 80K 100K SE +/- 232.81, N = 3 SE +/- 232.97, N = 3 SE +/- 87.23, N = 3 SE +/- 186.80, N = 3 SE +/- 198.41, N = 3 SE +/- 656.63, N = 3 SE +/- 242.45, N = 3 SE +/- 143.25, N = 3 SE +/- 78.02, N = 3 SE +/- 95.84, N = 3 SE +/- 137.82, N = 3 SE +/- 175.33, N = 3 SE +/- 434.11, N = 12 SE +/- 440.63, N = 3 SE +/- 58.85, N = 3 SE +/- 195.77, N = 3 SE +/- 865.71, N = 12 SE +/- 299.71, N = 3 SE +/- 1059.49, N = 12 SE +/- 221.82, N = 3 SE +/- 163.06, N = 3 SE +/- 106.16, N = 3 86677 79987 79492 79191 75396 73042 72244 72225 71227 69969 67315 61272 61233 60046 59228 58952 55505 55395 55272 38043 34716 33110 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
Blender Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.1 Blend File: Barbershop - Compute: CPU-Only EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 EPYC 9684X Xeon Platinum 8592+ 2P EPYC 9654 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E Xeon Platinum 8380 2P Xeon Platinum 8592+ EPYC 8534P Xeon Platinum 8490H EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 100 200 300 400 500 SE +/- 0.07, N = 3 SE +/- 0.11, N = 3 SE +/- 0.02, N = 3 SE +/- 0.17, N = 3 SE +/- 0.41, N = 3 SE +/- 0.03, N = 3 SE +/- 0.12, N = 3 SE +/- 0.21, N = 3 SE +/- 0.25, N = 3 SE +/- 0.42, N = 3 SE +/- 0.06, N = 3 SE +/- 0.23, N = 3 SE +/- 0.08, N = 3 SE +/- 0.33, N = 3 SE +/- 0.25, N = 3 SE +/- 0.69, N = 3 SE +/- 0.06, N = 3 SE +/- 0.18, N = 3 SE +/- 0.07, N = 3 SE +/- 1.09, N = 3 SE +/- 0.12, N = 3 SE +/- 0.03, N = 3 65.58 67.52 79.05 97.18 106.48 114.76 123.90 125.87 139.25 141.54 156.70 169.44 174.49 191.20 235.31 236.73 240.13 261.97 263.76 290.56 320.11 456.81 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Blender 4.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9654 EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.54 / Avg: 139.56 / Max: 146.03 Min: 17.05 / Avg: 166.92 / Max: 172.9 Min: 43.74 / Avg: 219.92 / Max: 232.7 Min: 52.6 / Avg: 265.95 / Max: 290.85 Min: 43.73 / Avg: 282.74 / Max: 303.29 Min: 27.59 / Avg: 319.36 / Max: 343.48 Min: 23.25 / Avg: 321.43 / Max: 348.37 Min: 33.95 / Avg: 339.43 / Max: 377.51 Min: 60.47 / Avg: 340.87 / Max: 376.73 Min: 85.64 / Avg: 342.89 / Max: 377.27 Min: 73.43 / Avg: 343.06 / Max: 375.17 Min: 35.25 / Avg: 388.09 / Max: 411.05 Min: 106.4 / Avg: 442.07 / Max: 501.67 Min: 108.93 / Avg: 517.76 / Max: 626.1 Min: 104.16 / Avg: 521 / Max: 582.69 Min: 41.32 / Avg: 614.31 / Max: 703.89 Min: 58.63 / Avg: 632.52 / Max: 694.54 Min: 79.57 / Avg: 660.14 / Max: 755.17 Min: 167.6 / Avg: 666.99 / Max: 755.27 Min: 209.81 / Avg: 673.68 / Max: 752.4 Min: 206.17 / Avg: 675.87 / Max: 751.49 Min: 68.76 / Avg: 738.74 / Max: 823.72
Numpy Benchmark This is a test to obtain the general Numpy performance. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Score, More Is Better Numpy Benchmark EPYC 9684X EPYC 9684X 2P EPYC 9654 EPYC 9654 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 120 240 360 480 600 SE +/- 1.85, N = 3 SE +/- 0.60, N = 3 SE +/- 2.28, N = 3 SE +/- 0.38, N = 3 SE +/- 1.84, N = 3 SE +/- 1.42, N = 3 SE +/- 0.79, N = 3 SE +/- 0.60, N = 3 SE +/- 0.26, N = 3 SE +/- 0.33, N = 3 SE +/- 1.30, N = 3 SE +/- 1.13, N = 3 SE +/- 0.96, N = 3 SE +/- 0.25, N = 3 SE +/- 0.72, N = 3 SE +/- 1.34, N = 3 SE +/- 2.23, N = 3 SE +/- 1.52, N = 3 SE +/- 2.24, N = 3 SE +/- 0.47, N = 3 SE +/- 0.37, N = 3 SE +/- 0.30, N = 3 542.15 540.96 529.80 529.03 520.22 519.89 464.72 461.61 456.18 455.57 439.84 435.21 425.46 422.76 421.99 418.89 400.64 394.71 284.33 283.52 261.19 258.49 Score Per Watt
OpenBenchmarking.org Score Per Watt, More Is Better Numpy Benchmark EPYC 8534PN EPYC 8534P Xeon 6780E Xeon 6766E EPYC 9754 EPYC 9654 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9754 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 2 4 6 8 10 7.962 7.918 6.244 5.817 5.405 5.359 5.255 4.854 4.183 4.043 3.290 3.164 2.603 2.382 2.360 2.249 2.170 2.079 2.071 1.716 1.375 1.343 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Numpy Benchmark CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8490H EPYC 9684X Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 80 160 240 320 400 Min: 21.86 / Avg: 44.9 / Max: 66.76 Min: 43.42 / Avg: 45.54 / Max: 66.48 Min: 14.57 / Avg: 52.61 / Max: 56.75 Min: 16.02 / Avg: 53.3 / Max: 57.51 Min: 20.13 / Avg: 78.72 / Max: 83.54 Min: 26.33 / Avg: 98.86 / Max: 103.35 Min: 33.91 / Avg: 98.99 / Max: 138.69 Min: 48.6 / Avg: 99.09 / Max: 103.63 Min: 104.03 / Avg: 108.54 / Max: 129.89 Min: 53.93 / Avg: 108.91 / Max: 127.19 Min: 59.12 / Avg: 108.92 / Max: 159.89 Min: 32.28 / Avg: 111.7 / Max: 116.99 Min: 73.46 / Avg: 140.29 / Max: 199.25 Min: 82.95 / Avg: 146.89 / Max: 202.52 Min: 99.11 / Avg: 190.58 / Max: 201.58 Min: 38.63 / Avg: 194.78 / Max: 202.23 Min: 78.29 / Avg: 220.31 / Max: 293.25 Min: 57.9 / Avg: 235.19 / Max: 242.43 Min: 64.52 / Avg: 260.26 / Max: 267.94 Min: 168.75 / Avg: 265.77 / Max: 348.44 Min: 203.89 / Avg: 319.82 / Max: 421.94 Min: 137.87 / Avg: 324.03 / Max: 425.17
LuxCoreRender LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org M samples/sec, More Is Better LuxCoreRender 2.6 Scene: DLSC - Acceleration: CPU EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P EPYC 9754 EPYC 9684X EPYC 9654 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E EPYC 8534P Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H EPYC 8534PN Xeon Max 9480 Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9468 Xeon Platinum 8380 5 10 15 20 25 SE +/- 0.91, N = 15 SE +/- 1.02, N = 15 SE +/- 0.12, N = 3 SE +/- 0.28, N = 15 SE +/- 0.18, N = 15 SE +/- 0.13, N = 3 SE +/- 0.24, N = 15 SE +/- 0.23, N = 15 SE +/- 0.20, N = 15 SE +/- 0.12, N = 3 SE +/- 0.06, N = 3 SE +/- 0.02, N = 3 SE +/- 0.06, N = 15 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 SE +/- 0.06, N = 15 SE +/- 0.03, N = 3 SE +/- 0.10, N = 3 SE +/- 0.22, N = 15 SE +/- 0.22, N = 12 SE +/- 0.08, N = 3 SE +/- 0.05, N = 3 22.58 22.00 17.89 16.79 16.21 14.29 12.74 11.12 10.44 10.17 10.09 9.40 9.39 9.38 8.98 8.72 8.48 8.00 7.99 7.71 7.46 5.63 M samples/sec Per Watt
OpenBenchmarking.org M samples/sec Per Watt, More Is Better LuxCoreRender 2.6 Scene: DLSC - Acceleration: CPU EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6766E EPYC 9654 Xeon 6780E EPYC 9684X EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ Xeon 6766E 2P Xeon Platinum 8490H Xeon Max 9480 Xeon 6780E 2P Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.0149 0.0298 0.0447 0.0596 0.0745 0.066 0.060 0.054 0.052 0.048 0.045 0.042 0.039 0.031 0.031 0.029 0.026 0.026 0.024 0.023 0.023 0.022 0.019 0.018 0.017 0.016 0.015 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better LuxCoreRender 2.6 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 Xeon 6766E 2P EPYC 9654 EPYC 9754 Xeon Platinum 8592+ Xeon Max 9480 Xeon Platinum 8490H Xeon Max 9468 Xeon 6780E 2P EPYC 9684X Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 150 300 450 600 750 Min: 18.19 / Avg: 128.95 / Max: 140.58 Min: 18.62 / Avg: 156.14 / Max: 171.55 Min: 43.65 / Avg: 180.3 / Max: 223.44 Min: 43.51 / Avg: 225.44 / Max: 288.32 Min: 51.32 / Avg: 255.3 / Max: 289.97 Min: 104.02 / Avg: 295.99 / Max: 409.35 Min: 29.9 / Avg: 299.39 / Max: 326.71 Min: 20.73 / Avg: 311.89 / Max: 340.37 Min: 24.28 / Avg: 326.43 / Max: 375.69 Min: 72.86 / Avg: 330.29 / Max: 377.65 Min: 59.67 / Avg: 330.53 / Max: 373.83 Min: 74.56 / Avg: 331.26 / Max: 374.04 Min: 79.88 / Avg: 341.83 / Max: 482.57 Min: 34.75 / Avg: 384.12 / Max: 412.45 Min: 101.11 / Avg: 511.71 / Max: 580 Min: 42.95 / Avg: 561.1 / Max: 675.69 Min: 59.66 / Avg: 573.93 / Max: 661.76 Min: 76.68 / Avg: 654.92 / Max: 755.35 Min: 173.98 / Avg: 664.56 / Max: 753.01 Min: 195.7 / Avg: 665.76 / Max: 753.2 Min: 203.55 / Avg: 668.4 / Max: 754.98 Min: 65.89 / Avg: 739.45 / Max: 837.69
OpenSSL OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org byte/s, More Is Better OpenSSL 3.3 Algorithm: AES-128-GCM EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E 2P EPYC 9754 Xeon 6766E 2P EPYC 9684X EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6780E Xeon 6766E EPYC 8534P EPYC 8534PN Xeon Platinum 8380 500000M 1000000M 1500000M 2000000M 2500000M SE +/- 1841260363.24, N = 3 SE +/- 1428966819.54, N = 3 SE +/- 337702856.52, N = 3 SE +/- 1622921637.03, N = 3 SE +/- 159598753.20, N = 3 SE +/- 697208281.49, N = 3 SE +/- 268465891.11, N = 3 SE +/- 333650111.34, N = 3 SE +/- 49279694.68, N = 3 SE +/- 429428665.34, N = 3 SE +/- 570058274.06, N = 3 SE +/- 164763342.54, N = 3 SE +/- 1253503549.61, N = 3 SE +/- 266669781.18, N = 3 SE +/- 28627701.54, N = 3 SE +/- 56355870.99, N = 3 SE +/- 43545680.66, N = 3 SE +/- 129279575.06, N = 3 SE +/- 195953395.92, N = 3 SE +/- 530918594.60, N = 3 SE +/- 409615403.40, N = 3 SE +/- 706172241.68, N = 3 2316611945283 1990352059967 1785006879067 1769313146013 1533026300727 1484669627050 1277583382757 1266882866653 1167221141870 1096082199713 996007109267 897288436123 889567506540 807229265300 768804693697 747710702660 641200848170 638453203763 565720767783 539784931373 488965390857 399526815427 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl byte/s Per Watt
OpenBenchmarking.org byte/s Per Watt, More Is Better OpenSSL 3.3 Algorithm: AES-128-GCM EPYC 8534PN EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 9654 EPYC 9654 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9684X 2P Xeon Platinum 8490H Xeon 6766E Xeon Platinum 8490H 2P Xeon 6766E 2P Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon 6780E Xeon Max 9468 2P Xeon 6780E 2P Xeon Platinum 8380 2P Xeon Platinum 8380 700M 1400M 2100M 2800M 3500M 3483917754.10 3351926704.37 3291374680.15 3225126995.22 2686039538.13 2638396952.00 2600338214.73 2589909503.87 2482510440.03 2480518885.86 2333237948.27 2305290901.87 2275619248.61 2230951841.30 2214398499.06 2168460340.43 2035473064.32 1970077384.37 1967868118.18 1953802430.20 1525932773.14 1510207805.12 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon Max 9468 Xeon 6780E Xeon Platinum 8490H EPYC 9654 Xeon Max 9480 Xeon Platinum 8592+ EPYC 9754 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P EPYC 9754 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.62 / Avg: 140.35 / Max: 146.9 Min: 17.13 / Avg: 167.37 / Max: 173.46 Min: 43.71 / Avg: 245.4 / Max: 270.99 Min: 54.11 / Avg: 264.55 / Max: 290.68 Min: 72.06 / Avg: 315.01 / Max: 350.37 Min: 43.57 / Avg: 324.08 / Max: 356.29 Min: 57.48 / Avg: 329.5 / Max: 379.19 Min: 28.74 / Avg: 334.06 / Max: 354.77 Min: 73.91 / Avg: 337.66 / Max: 377.26 Min: 21.98 / Avg: 342.1 / Max: 376.9 Min: 22.8 / Avg: 348.22 / Max: 361.46 Min: 34.83 / Avg: 401.21 / Max: 411.04 Min: 109.26 / Avg: 491.31 / Max: 541.12 Min: 101.06 / Avg: 529.01 / Max: 581.52 Min: 110.15 / Avg: 648.42 / Max: 712.91 Min: 201.68 / Avg: 649.22 / Max: 717.93 Min: 164.19 / Avg: 673.67 / Max: 754.31 Min: 56.69 / Avg: 676.55 / Max: 719.01 Min: 68.61 / Avg: 683.16 / Max: 735.37 Min: 187.64 / Avg: 684.67 / Max: 756.52 Min: 40.47 / Avg: 703.84 / Max: 726.54 Min: 68.36 / Avg: 802.39 / Max: 822.1
Result
OpenBenchmarking.org byte/s, More Is Better OpenSSL 3.3 Algorithm: AES-256-GCM EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon 6780E 2P Xeon Max 9468 2P EPYC 9754 Xeon 6766E 2P EPYC 9684X EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon 6780E Xeon Max 9468 Xeon 6766E EPYC 8534P EPYC 8534PN Xeon Platinum 8380 400000M 800000M 1200000M 1600000M 2000000M SE +/- 1791964708.17, N = 3 SE +/- 1859275059.09, N = 3 SE +/- 2225250504.62, N = 3 SE +/- 1320809584.03, N = 3 SE +/- 1397764228.85, N = 3 SE +/- 112278705.55, N = 3 SE +/- 301551944.80, N = 3 SE +/- 108288170.03, N = 3 SE +/- 290015345.25, N = 3 SE +/- 252081360.30, N = 3 SE +/- 671752057.50, N = 3 SE +/- 818689705.82, N = 3 SE +/- 3966858436.56, N = 3 SE +/- 116420153.53, N = 3 SE +/- 2848140790.67, N = 3 SE +/- 63026881.96, N = 3 SE +/- 317370827.95, N = 3 SE +/- 119857064.56, N = 3 SE +/- 34244051.67, N = 3 SE +/- 148092414.72, N = 3 SE +/- 3442274884.66, N = 3 SE +/- 81945775.93, N = 3 1983103026700 1714061761087 1526505465813 1497529939377 1369534328470 1181148957353 1125373041173 1011898993323 997539007873 979419388797 859720599963 767696553667 744387683670 712485411447 700239352717 592014115453 568937578313 507456728087 506301374623 462309750530 415893963207 356062040883 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl byte/s Per Watt
OpenBenchmarking.org byte/s Per Watt, More Is Better OpenSSL 3.3 Algorithm: AES-256-GCM EPYC 8534PN EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 9654 EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8592+ EPYC 9684X EPYC 9684X 2P Xeon 6766E Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon 6766E 2P Xeon Max 9480 Xeon 6780E Xeon Max 9480 2P Xeon 6780E 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Platinum 8380 600M 1200M 1800M 2400M 3000M 2991193842.10 2889219779.92 2844279712.45 2800094011.98 2331206946.71 2286984508.79 2193171119.61 2177657519.61 2144285710.42 2139406659.57 2061752146.93 2048623453.07 1998240264.78 1993158291.38 1789501537.54 1758065363.32 1745319585.66 1735227671.90 1651813451.15 1587520126.44 1345006115.81 1344485766.05 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon Max 9468 Xeon 6780E EPYC 9654 Xeon Max 9480 Xeon Platinum 8490H Xeon Platinum 8592+ EPYC 9754 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon 6780E 2P EPYC 9654 2P Xeon Max 9480 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9754 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.68 / Avg: 139.04 / Max: 145.32 Min: 17.84 / Avg: 165.11 / Max: 170.43 Min: 43.78 / Avg: 245.57 / Max: 271.22 Min: 52.04 / Avg: 264.83 / Max: 290.23 Min: 69.15 / Avg: 307.21 / Max: 344.49 Min: 43.56 / Avg: 323.62 / Max: 356.6 Min: 29.29 / Avg: 329.31 / Max: 350.94 Min: 76.66 / Avg: 330.83 / Max: 377.06 Min: 57.11 / Avg: 341.81 / Max: 377.46 Min: 23.55 / Avg: 341.83 / Max: 377.46 Min: 22.57 / Avg: 345.26 / Max: 357.86 Min: 34.11 / Avg: 400.94 / Max: 411.04 Min: 108.93 / Avg: 491.39 / Max: 542.48 Min: 103.41 / Avg: 529.73 / Max: 581.31 Min: 186.02 / Avg: 637.41 / Max: 707.3 Min: 109.57 / Avg: 648.54 / Max: 712.9 Min: 59.32 / Avg: 667.48 / Max: 710.25 Min: 191.65 / Avg: 676.75 / Max: 755.44 Min: 67.1 / Avg: 682.81 / Max: 732.1 Min: 152.63 / Avg: 685.37 / Max: 754.67 Min: 42.68 / Avg: 697.23 / Max: 724.85 Min: 64.77 / Avg: 801.19 / Max: 822.22
Result
OpenBenchmarking.org byte/s, More Is Better OpenSSL 3.3 Algorithm: ChaCha20 EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9754 Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9480 2P EPYC 9654 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 8534P EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 Xeon 6780E Xeon 6766E 300000M 600000M 900000M 1200000M 1500000M SE +/- 63368290.15, N = 3 SE +/- 874940734.18, N = 3 SE +/- 80233613.02, N = 3 SE +/- 303075210.03, N = 3 SE +/- 170968063.89, N = 3 SE +/- 271318554.29, N = 3 SE +/- 94935214.36, N = 3 SE +/- 863978337.15, N = 3 SE +/- 151272595.84, N = 3 SE +/- 329510485.23, N = 3 SE +/- 352722951.98, N = 3 SE +/- 63366974.35, N = 3 SE +/- 11835135.62, N = 3 SE +/- 3335467.26, N = 3 SE +/- 232288636.76, N = 3 SE +/- 9570566.28, N = 3 SE +/- 71110585.03, N = 3 SE +/- 312824840.64, N = 3 SE +/- 415487944.57, N = 3 SE +/- 151152302.42, N = 3 SE +/- 39095934.62, N = 3 SE +/- 11028602.79, N = 3 1306928507573 1086618191950 1001502634860 684284382760 656396329387 624402972947 544904084773 522912403347 503311021123 449908494927 425451190063 391911225480 352826821653 344150916937 316126421173 307068938377 282053327077 263740690067 226597354153 212294883830 196920683270 177249062183 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl byte/s Per Watt
OpenBenchmarking.org byte/s Per Watt, More Is Better OpenSSL 3.3 Algorithm: ChaCha20 EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9754 2P EPYC 9654 EPYC 9654 2P EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon 6766E Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Platinum 8380 2P Xeon 6766E 2P Xeon Platinum 8380 Xeon Max 9468 2P Xeon 6780E Xeon 6780E 2P 400M 800M 1200M 1600M 2000M 2009755510.33 1824186596.16 1795036060.88 1778324265.62 1510491450.86 1480134319.02 1361592598.33 1353849347.71 1008751851.93 1003586866.34 977212174.27 936302176.04 862670148.37 852735838.29 846331621.99 811931853.54 804835329.47 804207062.62 802490465.58 774438543.71 730208953.76 712317847.22 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H EPYC 9654 Xeon Platinum 8592+ EPYC 9754 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9754 2P EPYC 9684X 2P 150 300 450 600 750 Min: 15.27 / Avg: 140.34 / Max: 147.74 Min: 17.59 / Avg: 168.33 / Max: 176.1 Min: 43.7 / Avg: 205.47 / Max: 234.45 Min: 54.01 / Avg: 264.55 / Max: 291.24 Min: 43.55 / Avg: 269.68 / Max: 311.17 Min: 71.67 / Avg: 279.08 / Max: 326.73 Min: 73.21 / Avg: 309.29 / Max: 373.85 Min: 54.19 / Avg: 323.5 / Max: 377.22 Min: 30.31 / Avg: 333.21 / Max: 358.89 Min: 22.69 / Avg: 341.17 / Max: 376.63 Min: 22.28 / Avg: 365.67 / Max: 391.73 Min: 34.55 / Avg: 400.2 / Max: 411.05 Min: 108.7 / Avg: 438.73 / Max: 492.61 Min: 101.74 / Avg: 528.62 / Max: 581.46 Min: 109.96 / Avg: 550.19 / Max: 629.46 Min: 197.94 / Avg: 580.95 / Max: 674.82 Min: 183.97 / Avg: 617.86 / Max: 744.83 Min: 168.89 / Avg: 666.88 / Max: 755.69 Min: 59.94 / Avg: 676.63 / Max: 728.84 Min: 70.13 / Avg: 681.84 / Max: 728.96 Min: 42.15 / Avg: 734.92 / Max: 800.62 Min: 70.23 / Avg: 802.61 / Max: 826.12
Result
OpenBenchmarking.org byte/s, More Is Better OpenSSL 3.3 Algorithm: SHA512 EPYC 9754 2P EPYC 9684X 2P Xeon 6780E 2P EPYC 9654 2P Xeon 6766E 2P EPYC 9754 Xeon Platinum 8592+ 2P EPYC 9684X Xeon 6780E Xeon Platinum 8490H 2P Xeon 6766E EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P EPYC 8534P Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 8534PN Xeon Platinum 8380 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 20000M 40000M 60000M 80000M 100000M SE +/- 17716195.81, N = 3 SE +/- 44023374.36, N = 3 SE +/- 64065176.93, N = 3 SE +/- 37307640.51, N = 3 SE +/- 66704467.17, N = 3 SE +/- 5822493.86, N = 3 SE +/- 14734422.48, N = 3 SE +/- 55712209.05, N = 3 SE +/- 24045021.87, N = 3 SE +/- 481040738.21, N = 3 SE +/- 14430784.68, N = 3 SE +/- 6285194.15, N = 3 SE +/- 14339684.98, N = 3 SE +/- 3500318.33, N = 3 SE +/- 15082903.45, N = 3 SE +/- 37654528.63, N = 3 SE +/- 203213936.42, N = 3 SE +/- 9990938.84, N = 3 SE +/- 20492408.59, N = 3 SE +/- 1250834.62, N = 3 SE +/- 500176.69, N = 3 SE +/- 6313493.56, N = 3 104776633047 91455971200 89194416857 78822257383 78599395230 52746634113 46520760707 45865131913 45088755640 42846254477 40605137923 39627575497 37564522053 32304222993 24378460820 23886054770 22219927723 22194470973 22079539900 18893980783 16199261277 10980940093 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl byte/s Per Watt
OpenBenchmarking.org byte/s Per Watt, More Is Better OpenSSL 3.3 Algorithm: SHA512 Xeon 6766E EPYC 8534PN Xeon 6766E 2P EPYC 9754 EPYC 9754 2P EPYC 8534P Xeon 6780E Xeon 6780E 2P EPYC 9654 EPYC 9654 2P EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Platinum 8380 40M 80M 120M 160M 200M 165922834.04 165874523.26 159806130.44 157697655.07 154893248.59 154425638.15 139323873.72 137868745.00 128564080.98 125714359.16 114623223.18 114225793.10 69884736.57 67966151.09 65144677.30 62457308.50 57121711.23 55449110.37 54000562.67 51580447.52 42041804.18 41470367.43 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon Max 9468 EPYC 9654 Xeon 6780E Xeon Max 9480 EPYC 9754 Xeon Platinum 8490H Xeon Platinum 8592+ EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon Max 9468 2P EPYC 9654 2P Xeon 6780E 2P EPYC 9754 2P Xeon Max 9480 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P 140 280 420 560 700 Min: 15.55 / Avg: 133.8 / Max: 143.63 Min: 18.44 / Avg: 157.87 / Max: 169.7 Min: 43.65 / Avg: 244.72 / Max: 270.07 Min: 53.35 / Avg: 264.79 / Max: 291.44 Min: 71.29 / Avg: 299.98 / Max: 317.48 Min: 28.99 / Avg: 308.23 / Max: 338.17 Min: 43.5 / Avg: 323.63 / Max: 356.46 Min: 74.41 / Avg: 330.77 / Max: 354.3 Min: 21.49 / Avg: 334.48 / Max: 349.67 Min: 56.54 / Avg: 341.09 / Max: 377.07 Min: 23.35 / Avg: 341.79 / Max: 377.28 Min: 34.64 / Avg: 400.14 / Max: 413.51 Min: 106.79 / Avg: 491.84 / Max: 540.07 Min: 105.66 / Avg: 525.18 / Max: 583.12 Min: 190.86 / Avg: 626.29 / Max: 729.82 Min: 59.93 / Avg: 626.99 / Max: 687.56 Min: 108.54 / Avg: 646.95 / Max: 712.56 Min: 42.68 / Avg: 676.44 / Max: 705.8 Min: 194.96 / Avg: 677.46 / Max: 722.73 Min: 66.13 / Avg: 684.47 / Max: 754.35 Min: 158.63 / Avg: 686.01 / Max: 753.85 Min: 66.79 / Avg: 800.66 / Max: 822.16
Result
OpenBenchmarking.org byte/s, More Is Better OpenSSL 3.3 Algorithm: ChaCha20-Poly1305 EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P EPYC 9754 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9480 2P EPYC 9654 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8490H Xeon Platinum 8592+ EPYC 8534P EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 Xeon 6780E Xeon 6766E 200000M 400000M 600000M 800000M 1000000M SE +/- 19082587.00, N = 3 SE +/- 558033954.71, N = 3 SE +/- 6119305.11, N = 3 SE +/- 151303743.70, N = 3 SE +/- 82161991.11, N = 3 SE +/- 132227134.30, N = 3 SE +/- 4478378.70, N = 3 SE +/- 36941390.18, N = 3 SE +/- 29295317.86, N = 3 SE +/- 4148679.28, N = 3 SE +/- 31274617.76, N = 3 SE +/- 38565703.74, N = 3 SE +/- 37182497.08, N = 3 SE +/- 766713911.50, N = 3 SE +/- 120899798.30, N = 3 SE +/- 24234601.89, N = 3 SE +/- 65233629.47, N = 3 SE +/- 37795783.50, N = 3 SE +/- 6370607.14, N = 3 SE +/- 255765711.98, N = 3 SE +/- 18243420.30, N = 3 SE +/- 2360978.39, N = 3 906028608990 767652719547 698925822223 457039748870 444099144270 442202532977 383925652077 372735215297 350960998357 320396720447 293502122413 266583857887 240005135153 225038850730 223625738013 211185661727 193305093643 187796174440 161321058213 146156212950 134033039090 120591787257 1. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl byte/s Per Watt
OpenBenchmarking.org byte/s Per Watt, More Is Better OpenSSL 3.3 Algorithm: ChaCha20-Poly1305 EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9754 2P EPYC 9654 EPYC 9654 2P EPYC 9684X 2P EPYC 9684X Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9480 2P Xeon 6766E Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon Max 9468 2P Xeon 6766E 2P Xeon 6780E Xeon 6780E 2P 300M 600M 900M 1200M 1500M 1388893901.10 1287300050.02 1269363721.24 1244610923.33 1061980440.86 1039886433.31 957228032.13 957073830.82 666607882.23 658820863.84 648595124.99 644676877.17 603664493.39 574875604.57 569925760.92 568545794.25 554436256.62 552257367.63 542055990.77 536427227.19 483646931.54 471667041.65 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenSSL 3.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E Xeon Max 9468 Xeon Max 9480 EPYC 9654 Xeon Platinum 8490H Xeon Platinum 8592+ EPYC 9754 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9754 2P EPYC 9684X 2P 140 280 420 560 700 Min: 14.85 / Avg: 139.18 / Max: 148.57 Min: 16.74 / Avg: 164.05 / Max: 171.53 Min: 43.7 / Avg: 211.59 / Max: 250.8 Min: 54.07 / Avg: 264.65 / Max: 291.12 Min: 43.58 / Avg: 277.13 / Max: 331.03 Min: 74.91 / Avg: 283.74 / Max: 304.03 Min: 71.53 / Avg: 311.09 / Max: 362.15 Min: 27.82 / Avg: 330.48 / Max: 353.86 Min: 58.01 / Avg: 337.59 / Max: 352.56 Min: 21.83 / Avg: 339.43 / Max: 376.89 Min: 21.97 / Avg: 360.05 / Max: 375.06 Min: 35.8 / Avg: 401.15 / Max: 411.05 Min: 107.32 / Avg: 447.17 / Max: 525.92 Min: 104.8 / Avg: 529.37 / Max: 582.49 Min: 109.07 / Avg: 565.2 / Max: 671.32 Min: 186.72 / Avg: 591.08 / Max: 630.87 Min: 193.92 / Avg: 648.38 / Max: 695.92 Min: 54.76 / Avg: 672.12 / Max: 721.51 Min: 79.72 / Avg: 684.71 / Max: 754.83 Min: 163.33 / Avg: 685.93 / Max: 736.16 Min: 41.79 / Avg: 727.96 / Max: 758.79 Min: 68.91 / Avg: 801.95 / Max: 822.13
PyTorch This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org batches/sec, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 512 - Model: ResNet-50 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 2P EPYC 9684X EPYC 9654 EPYC 8534P EPYC 8534PN Xeon 6780E EPYC 9754 Xeon 6766E EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P 14 28 42 56 70 SE +/- 0.19, N = 3 SE +/- 0.12, N = 3 SE +/- 0.43, N = 15 SE +/- 0.72, N = 3 SE +/- 0.17, N = 3 SE +/- 0.26, N = 3 SE +/- 0.30, N = 15 SE +/- 0.34, N = 15 SE +/- 0.21, N = 3 SE +/- 0.38, N = 15 SE +/- 0.19, N = 3 SE +/- 0.10, N = 3 SE +/- 0.08, N = 3 SE +/- 0.11, N = 3 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 SE +/- 0.24, N = 3 SE +/- 0.29, N = 12 SE +/- 0.23, N = 3 SE +/- 0.13, N = 9 SE +/- 0.03, N = 3 SE +/- 0.22, N = 12 63.30 56.44 51.86 51.05 48.99 41.65 40.83 37.84 37.55 37.45 34.69 34.65 33.76 33.18 30.87 27.53 26.91 16.82 16.24 15.97 14.81 13.76 batches/sec Per Watt
OpenBenchmarking.org batches/sec Per Watt, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 512 - Model: ResNet-50 EPYC 8534PN EPYC 8534P Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H Xeon 6780E Xeon 6766E Xeon Max 9468 Xeon Max 9480 EPYC 9654 EPYC 9754 EPYC 9684X Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9654 2P EPYC 9684X 2P 0.0747 0.1494 0.2241 0.2988 0.3735 0.332 0.305 0.258 0.236 0.225 0.194 0.194 0.192 0.188 0.178 0.175 0.149 0.114 0.080 0.077 0.073 0.066 0.046 0.045 0.045 0.044 0.040 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better PyTorch 2.2.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon Platinum 8380 2P Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 110 220 330 440 550 Min: 17.13 / Avg: 99.8 / Max: 111.33 Min: 18.44 / Avg: 110.66 / Max: 125.07 Min: 43.57 / Avg: 138.61 / Max: 153.19 Min: 21.06 / Avg: 157.61 / Max: 175.55 Min: 43.48 / Avg: 159.18 / Max: 178.15 Min: 30.14 / Avg: 194.55 / Max: 219.58 Min: 56.55 / Avg: 207.89 / Max: 238.03 Min: 35.86 / Avg: 232.68 / Max: 261.18 Min: 34.92 / Avg: 245.26 / Max: 293.93 Min: 59.74 / Avg: 250.75 / Max: 298.2 Min: 74.44 / Avg: 265.6 / Max: 309.82 Min: 70.81 / Avg: 276.02 / Max: 323.04 Min: 39.07 / Avg: 298.03 / Max: 341.96 Min: 98.32 / Avg: 327.4 / Max: 419.31 Min: 105.19 / Avg: 327.79 / Max: 351.26 Min: 108.16 / Avg: 354.05 / Max: 391.41 Min: 58.93 / Avg: 369.53 / Max: 409.87 Min: 63.73 / Avg: 418.64 / Max: 475.03 Min: 161.18 / Avg: 490.32 / Max: 568.06 Min: 79.08 / Avg: 511.14 / Max: 575.41 Min: 202.76 / Avg: 570.95 / Max: 625.38 Min: 176.16 / Avg: 577.13 / Max: 637.06
LuxCoreRender LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org M samples/sec, More Is Better LuxCoreRender 2.6 Scene: LuxCore Benchmark - Acceleration: CPU EPYC 9684X EPYC 9754 EPYC 9654 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon 6780E Xeon Platinum 8592+ Xeon 6766E Xeon Max 9480 2P EPYC 8534P Xeon Platinum 8490H EPYC 8534PN Xeon Max 9480 Xeon Max 9468 2P Xeon 6780E 2P Xeon Platinum 8380 2P Xeon 6766E 2P Xeon Max 9468 Xeon Platinum 8380 3 6 9 12 15 SE +/- 0.04, N = 3 SE +/- 0.16, N = 3 SE +/- 0.10, N = 7 SE +/- 0.12, N = 15 SE +/- 0.11, N = 15 SE +/- 0.08, N = 15 SE +/- 0.17, N = 12 SE +/- 0.12, N = 3 SE +/- 0.06, N = 3 SE +/- 0.11, N = 3 SE +/- 0.06, N = 3 SE +/- 0.16, N = 15 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 SE +/- 0.22, N = 12 SE +/- 0.08, N = 15 SE +/- 0.07, N = 15 SE +/- 0.10, N = 15 SE +/- 0.03, N = 3 SE +/- 0.06, N = 15 12.43 12.41 11.34 10.61 10.56 9.65 8.88 8.82 8.47 8.06 7.90 7.41 7.38 7.10 6.93 6.82 6.81 6.74 6.43 6.41 6.30 4.29 M samples/sec Per Watt
OpenBenchmarking.org M samples/sec Per Watt, More Is Better LuxCoreRender 2.6 Scene: LuxCore Benchmark - Acceleration: CPU EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 EPYC 9654 Xeon 6780E EPYC 9684X Xeon Platinum 8592+ Xeon 6766E 2P Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9480 Xeon 6780E 2P EPYC 9654 2P Xeon Max 9468 EPYC 9684X 2P Xeon Platinum 8380 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.0119 0.0238 0.0357 0.0476 0.0595 0.053 0.049 0.046 0.043 0.040 0.039 0.034 0.025 0.022 0.022 0.021 0.021 0.020 0.020 0.019 0.017 0.017 0.014 0.014 0.013 0.011 0.010 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better LuxCoreRender 2.6 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9654 EPYC 9754 Xeon 6766E 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6780E 2P EPYC 9684X EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 17.5 / Avg: 131.25 / Max: 148.58 Min: 18.49 / Avg: 149.9 / Max: 169.31 Min: 43.63 / Avg: 172.93 / Max: 201.7 Min: 43.54 / Avg: 215.89 / Max: 253.56 Min: 52.8 / Avg: 250 / Max: 286.38 Min: 28.93 / Avg: 284.2 / Max: 318.53 Min: 23.22 / Avg: 288.11 / Max: 334.63 Min: 104.76 / Avg: 297.7 / Max: 400.24 Min: 34.18 / Avg: 318.84 / Max: 374.24 Min: 62.21 / Avg: 319.52 / Max: 375.32 Min: 74.54 / Avg: 324.24 / Max: 366.89 Min: 85.32 / Avg: 325.06 / Max: 364.85 Min: 107.86 / Avg: 344.5 / Max: 460.96 Min: 34.2 / Avg: 365.21 / Max: 410.93 Min: 40.41 / Avg: 450.3 / Max: 521.1 Min: 101 / Avg: 492.26 / Max: 580.68 Min: 58.25 / Avg: 537.06 / Max: 610.46 Min: 38.08 / Avg: 620.51 / Max: 723.33 Min: 83.65 / Avg: 634.64 / Max: 754.69 Min: 178.58 / Avg: 636.83 / Max: 753.39 Min: 129.79 / Avg: 649.71 / Max: 752.19 Min: 205.6 / Avg: 650.43 / Max: 750.56
PyTorch This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org batches/sec, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 1 - Model: ResNet-152 Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P EPYC 9684X Xeon Platinum 8490H 2P EPYC 9654 EPYC 8534P EPYC 8534PN Xeon Max 9480 2P Xeon 6780E Xeon 6766E EPYC 9754 EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P EPYC 9754 2P Xeon 6766E 2P 6 12 18 24 30 SE +/- 0.08, N = 3 SE +/- 0.12, N = 3 SE +/- 0.26, N = 3 SE +/- 0.09, N = 3 SE +/- 0.16, N = 3 SE +/- 0.18, N = 3 SE +/- 0.19, N = 5 SE +/- 0.10, N = 3 SE +/- 0.01, N = 3 SE +/- 0.16, N = 4 SE +/- 0.09, N = 3 SE +/- 0.17, N = 3 SE +/- 0.06, N = 3 SE +/- 0.13, N = 8 SE +/- 0.06, N = 3 SE +/- 0.12, N = 3 SE +/- 0.04, N = 3 SE +/- 0.20, N = 12 SE +/- 0.27, N = 9 SE +/- 0.07, N = 3 SE +/- 0.15, N = 12 SE +/- 0.17, N = 9 27.21 22.27 19.99 19.41 18.07 17.91 17.57 16.68 15.86 15.77 15.57 15.25 15.11 14.86 14.53 13.39 13.05 8.35 8.27 7.10 6.64 6.47 batches/sec Per Watt
OpenBenchmarking.org batches/sec Per Watt, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 1 - Model: ResNet-152 EPYC 8534PN EPYC 8534P Xeon Platinum 8592+ Xeon Platinum 8380 Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8490H Xeon Max 9468 EPYC 9684X Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9754 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9684X 2P 0.0351 0.0702 0.1053 0.1404 0.1755 0.156 0.146 0.113 0.111 0.096 0.092 0.084 0.081 0.080 0.076 0.069 0.067 0.056 0.034 0.033 0.031 0.026 0.022 0.022 0.020 0.020 0.020 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better PyTorch 2.2.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9754 2P Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 110 220 330 440 550 Min: 16.2 / Avg: 97.09 / Max: 104.85 Min: 16.21 / Avg: 104.37 / Max: 113.9 Min: 43.54 / Avg: 139.02 / Max: 150.49 Min: 19.56 / Avg: 155.97 / Max: 167.94 Min: 43.41 / Avg: 158.21 / Max: 172.64 Min: 27.2 / Avg: 192.05 / Max: 207.21 Min: 55.11 / Avg: 200.49 / Max: 223.6 Min: 30.38 / Avg: 228.52 / Max: 247.96 Min: 33.7 / Avg: 239.78 / Max: 273.27 Min: 58.34 / Avg: 242.81 / Max: 264.97 Min: 73.95 / Avg: 262.34 / Max: 291.94 Min: 73.16 / Avg: 270.76 / Max: 301.09 Min: 102.12 / Avg: 295.35 / Max: 320.3 Min: 20.22 / Avg: 296.81 / Max: 327.42 Min: 54.23 / Avg: 317.38 / Max: 347.71 Min: 107.67 / Avg: 362.4 / Max: 382.35 Min: 55.84 / Avg: 369.03 / Max: 409.77 Min: 61.36 / Avg: 418.43 / Max: 475.94 Min: 162.11 / Avg: 476.49 / Max: 563.05 Min: 79.16 / Avg: 511.35 / Max: 559.55 Min: 196.88 / Avg: 571.95 / Max: 613.16 Min: 196.56 / Avg: 578.74 / Max: 617.41
Xmrig Xmrig is an open-source cross-platform CPU/GPU miner for RandomX, KawPow, CryptoNight and AstroBWT. This test profile is setup to measure the Xmrig CPU mining performance. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org H/s, More Is Better Xmrig 6.21 Variant: GhostRider - Hash Count: 1M EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon Max 9468 2P Xeon 6766E 2P EPYC 9754 Xeon Platinum 8592+ Xeon Max 9480 Xeon Platinum 8490H Xeon 6780E Xeon Max 9468 Xeon 6766E Xeon Platinum 8380 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8380 7K 14K 21K 28K 35K SE +/- 95.10, N = 3 SE +/- 290.50, N = 3 SE +/- 46.94, N = 3 SE +/- 64.69, N = 3 SE +/- 10.36, N = 3 SE +/- 117.32, N = 3 SE +/- 7.44, N = 3 SE +/- 4.03, N = 3 SE +/- 43.14, N = 3 SE +/- 14.02, N = 3 SE +/- 53.95, N = 3 SE +/- 61.27, N = 3 SE +/- 44.68, N = 3 SE +/- 14.78, N = 3 SE +/- 23.16, N = 3 SE +/- 43.00, N = 3 SE +/- 12.93, N = 3 SE +/- 19.74, N = 3 SE +/- 49.14, N = 12 SE +/- 30.53, N = 3 SE +/- 41.94, N = 3 SE +/- 32.20, N = 2 31157.8 26702.7 17873.1 16324.7 15786.6 14207.7 12115.0 11263.2 11147.2 10975.8 10431.6 9320.5 8378.3 6122.3 5808.8 5753.9 5446.8 5315.7 5230.9 4660.0 4564.3 2593.2 1. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc H/s Per Watt
OpenBenchmarking.org H/s Per Watt, More Is Better Xmrig 6.21 Variant: GhostRider - Hash Count: 1M EPYC 9654 EPYC 9654 2P EPYC 9684X 2P EPYC 9684X EPYC 9754 EPYC 8534PN EPYC 9754 2P EPYC 8534P Xeon 6766E Xeon 6780E Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Max 9480 2P Xeon Max 9480 Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Platinum 8380 11 22 33 44 55 46.688 45.051 44.082 42.004 40.166 36.236 36.168 33.466 30.621 29.190 28.504 27.676 25.291 25.015 18.002 17.929 17.114 16.828 16.636 16.293 10.183 9.798 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Xmrig 6.21 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon Max 9468 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 15.67 / Avg: 125.96 / Max: 131.33 Min: 17.62 / Avg: 139.24 / Max: 143.75 Min: 43.68 / Avg: 173.6 / Max: 178.87 Min: 43.43 / Avg: 197.12 / Max: 203.16 Min: 21.91 / Avg: 232.05 / Max: 250.74 Min: 53.3 / Avg: 264.66 / Max: 271.14 Min: 30.21 / Avg: 304.31 / Max: 328.13 Min: 73.18 / Avg: 327.41 / Max: 335.76 Min: 24.38 / Avg: 334.94 / Max: 357.54 Min: 55.89 / Avg: 339.42 / Max: 349.66 Min: 75.82 / Avg: 341.47 / Max: 356.09 Min: 103.58 / Avg: 365.97 / Max: 390.52 Min: 36 / Avg: 375.84 / Max: 410.13 Min: 104.75 / Avg: 402.77 / Max: 425.84 Min: 38.53 / Avg: 494.18 / Max: 551.71 Min: 95.12 / Avg: 513.7 / Max: 535.43 Min: 54.62 / Avg: 592.72 / Max: 673.62 Min: 73.19 / Avg: 645.48 / Max: 743.36 Min: 159.48 / Avg: 669.3 / Max: 730.91 Min: 188.92 / Avg: 672.99 / Max: 738.06 Min: 201.16 / Avg: 673.63 / Max: 702.68 Min: 60.4 / Avg: 706.81 / Max: 819.97
LuxCoreRender LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org M samples/sec, More Is Better LuxCoreRender 2.6 Scene: Orange Juice - Acceleration: CPU EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P EPYC 9684X EPYC 9754 EPYC 9654 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon Max 9468 2P Xeon 6766E 2P Xeon Max 9480 2P Xeon 6766E Xeon 6780E Xeon Platinum 8592+ Xeon Platinum 8380 2P EPYC 8534P Xeon Platinum 8490H EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 8 16 24 32 40 SE +/- 1.06, N = 15 SE +/- 1.12, N = 15 SE +/- 0.84, N = 15 SE +/- 0.28, N = 5 SE +/- 0.07, N = 3 SE +/- 0.18, N = 15 SE +/- 0.30, N = 15 SE +/- 0.12, N = 3 SE +/- 0.46, N = 15 SE +/- 0.16, N = 15 SE +/- 0.46, N = 12 SE +/- 0.28, N = 15 SE +/- 0.19, N = 3 SE +/- 0.06, N = 3 SE +/- 0.05, N = 3 SE +/- 0.06, N = 3 SE +/- 0.05, N = 3 SE +/- 0.07, N = 3 SE +/- 0.05, N = 3 SE +/- 0.04, N = 3 SE +/- 0.10, N = 3 SE +/- 0.03, N = 3 33.08 32.26 29.61 24.69 23.76 21.32 19.91 18.70 17.45 16.91 16.88 16.64 15.90 15.83 14.74 14.31 13.81 13.46 13.00 12.50 11.86 8.82 M samples/sec Per Watt
OpenBenchmarking.org M samples/sec Per Watt, More Is Better LuxCoreRender 2.6 Scene: Orange Juice - Acceleration: CPU EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 EPYC 9654 Xeon 6780E EPYC 9684X EPYC 9754 2P Xeon 6766E 2P EPYC 9654 2P Xeon 6780E 2P EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 0.0227 0.0454 0.0681 0.0908 0.1135 0.101 0.092 0.083 0.080 0.072 0.071 0.066 0.059 0.053 0.051 0.048 0.046 0.046 0.042 0.038 0.036 0.035 0.031 0.029 0.029 0.026 0.025 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better LuxCoreRender 2.6 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9654 EPYC 9754 Xeon Platinum 8592+ Xeon 6766E 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6780E 2P EPYC 9684X Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 150 300 450 600 750 Min: 17.62 / Avg: 128.09 / Max: 142.09 Min: 19.16 / Avg: 150.1 / Max: 167.85 Min: 43.68 / Avg: 191.46 / Max: 222.84 Min: 43.49 / Avg: 223.98 / Max: 283.31 Min: 53 / Avg: 249.69 / Max: 290.48 Min: 27.54 / Avg: 296.6 / Max: 330.72 Min: 22.04 / Avg: 298.61 / Max: 336.88 Min: 34.75 / Avg: 320.73 / Max: 377.48 Min: 104.86 / Avg: 321.3 / Max: 495.12 Min: 61.77 / Avg: 323.08 / Max: 377.72 Min: 75.77 / Avg: 326.79 / Max: 375.2 Min: 83.75 / Avg: 327.46 / Max: 377.5 Min: 107.87 / Avg: 364.17 / Max: 576.89 Min: 33.09 / Avg: 373.72 / Max: 411.13 Min: 100.16 / Avg: 499.14 / Max: 582.6 Min: 41.55 / Avg: 551.09 / Max: 677.8 Min: 57.15 / Avg: 586.21 / Max: 670.16 Min: 82.44 / Avg: 639.12 / Max: 751.58 Min: 172.79 / Avg: 646.27 / Max: 748.56 Min: 204.79 / Avg: 656.04 / Max: 751.44 Min: 201.92 / Avg: 656.25 / Max: 748.81 Min: 66.69 / Avg: 722.97 / Max: 825.83
OSPRay Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Items Per Second, More Is Better OSPRay 3.1 Benchmark: particle_volume/scivis/real_time EPYC 9754 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon 6780E 2P EPYC 9754 Xeon Max 9480 2P Xeon 6766E 2P EPYC 9684X Xeon Platinum 8592+ Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8490H Xeon 6780E Xeon 6766E Xeon Max 9480 EPYC 8534P Xeon Platinum 8380 2P EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 13 26 39 52 65 SE +/- 0.01106, N = 3 SE +/- 0.01792, N = 3 SE +/- 0.04682, N = 3 SE +/- 0.04960, N = 3 SE +/- 0.02499, N = 3 SE +/- 0.02131, N = 3 SE +/- 0.02252, N = 3 SE +/- 0.04134, N = 3 SE +/- 0.04097, N = 3 SE +/- 0.01858, N = 3 SE +/- 0.04057, N = 3 SE +/- 0.05872, N = 3 SE +/- 0.01265, N = 3 SE +/- 0.00525, N = 3 SE +/- 0.01460, N = 3 SE +/- 0.00610, N = 3 SE +/- 0.03593, N = 3 SE +/- 0.01395, N = 3 SE +/- 0.03711, N = 3 SE +/- 0.00583, N = 3 SE +/- 0.00932, N = 3 SE +/- 0.00909, N = 3 56.44850 55.25760 53.93480 47.52670 38.22690 36.24640 35.14640 33.12750 32.79790 32.52320 30.67990 29.34960 28.06890 21.89320 21.51470 19.42510 18.48510 17.94480 17.62220 16.45050 16.08690 9.70022 Items Per Second Per Watt
OpenBenchmarking.org Items Per Second Per Watt, More Is Better OSPRay 3.1 Benchmark: particle_volume/scivis/real_time EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9754 2P EPYC 9654 Xeon 6766E 2P Xeon Platinum 8592+ EPYC 9684X Xeon 6780E 2P Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 0.0326 0.0652 0.0978 0.1304 0.163 0.145 0.139 0.136 0.133 0.121 0.107 0.107 0.106 0.106 0.098 0.094 0.091 0.086 0.081 0.074 0.063 0.061 0.061 0.053 0.047 0.041 0.039 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OSPRay 3.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 Xeon Max 9468 EPYC 9654 EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.2 / Avg: 113.27 / Max: 134.96 Min: 18.42 / Avg: 129.01 / Max: 155.58 Min: 43.64 / Avg: 142.92 / Max: 193.36 Min: 43.6 / Avg: 178.41 / Max: 247.31 Min: 52.66 / Avg: 239.04 / Max: 286.42 Min: 74.87 / Avg: 255.3 / Max: 321.66 Min: 28.04 / Avg: 261.28 / Max: 302.78 Min: 22.25 / Avg: 264.01 / Max: 319.18 Min: 25.32 / Avg: 288.49 / Max: 368.86 Min: 59.59 / Avg: 296.81 / Max: 366.58 Min: 76.7 / Avg: 301.86 / Max: 366.25 Min: 105.71 / Avg: 310.52 / Max: 393.05 Min: 35.05 / Avg: 332.91 / Max: 399.84 Min: 108.67 / Avg: 386.65 / Max: 484.75 Min: 95.5 / Avg: 453.46 / Max: 555.04 Min: 42.57 / Avg: 526.9 / Max: 623.56 Min: 60.52 / Avg: 553.25 / Max: 625.16 Min: 67.98 / Avg: 591.05 / Max: 725.39 Min: 168.54 / Avg: 625.45 / Max: 728.13 Min: 191.04 / Avg: 627.62 / Max: 711.24 Min: 200.54 / Avg: 630.06 / Max: 725.99 Min: 67.89 / Avg: 682.63 / Max: 796.5
RocksDB This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Op/s, More Is Better RocksDB 9.0 Test: Read While Writing Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon 6780E Xeon 6766E Xeon Max 9480 2P Xeon Platinum 8380 2P Xeon Max 9468 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 6M 12M 18M 24M 30M SE +/- 726178.20, N = 12 SE +/- 426084.70, N = 15 SE +/- 273109.47, N = 3 SE +/- 506458.47, N = 15 SE +/- 236709.61, N = 3 SE +/- 109424.14, N = 3 SE +/- 36524.26, N = 3 SE +/- 93669.97, N = 15 SE +/- 215352.04, N = 15 SE +/- 343737.14, N = 15 SE +/- 136761.19, N = 4 SE +/- 45015.86, N = 3 SE +/- 254193.37, N = 15 SE +/- 113154.86, N = 4 SE +/- 218479.19, N = 15 SE +/- 28098.62, N = 3 SE +/- 81998.56, N = 3 SE +/- 28201.44, N = 3 SE +/- 61518.27, N = 10 SE +/- 83772.27, N = 13 SE +/- 26369.56, N = 3 SE +/- 23120.76, N = 3 27059786 24466143 22900152 20108289 19851664 15776186 14608544 14158419 12425519 12321300 11864815 11340389 11223329 9836689 9624417 9333645 8934176 8926893 7841738 7439663 7366500 5539870 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread Op/s Per Watt
OpenBenchmarking.org Op/s Per Watt, More Is Better RocksDB 9.0 Test: Read While Writing Xeon 6766E 2P Xeon 6780E 2P EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9754 2P EPYC 9684X EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 15K 30K 45K 60K 75K 69851.75 67516.29 66672.70 63183.03 60067.21 56900.62 52342.46 48605.69 45330.18 36906.83 34889.65 29572.89 27560.28 23723.76 22369.26 22256.95 22200.62 21648.99 19253.52 18781.27 16920.46 14775.49 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better RocksDB 9.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9654 EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.36 / Avg: 134 / Max: 149.2 Min: 17.34 / Avg: 155.39 / Max: 168.8 Min: 43.66 / Avg: 179.48 / Max: 270.38 Min: 43.57 / Avg: 208.52 / Max: 356.32 Min: 55.14 / Avg: 255.9 / Max: 290.56 Min: 24.6 / Avg: 300.55 / Max: 337.2 Min: 22 / Avg: 301.4 / Max: 339.84 Min: 33.11 / Avg: 323.9 / Max: 378.14 Min: 52.65 / Avg: 330.54 / Max: 376.66 Min: 74.89 / Avg: 331.82 / Max: 375.82 Min: 72.31 / Avg: 332.58 / Max: 376.04 Min: 107.56 / Avg: 350.26 / Max: 540.44 Min: 34.3 / Avg: 383.63 / Max: 411.02 Min: 108.91 / Avg: 400.79 / Max: 709.44 Min: 43.23 / Avg: 505.19 / Max: 682.09 Min: 98.57 / Avg: 510.9 / Max: 583.45 Min: 67.94 / Avg: 558.28 / Max: 752.93 Min: 56.63 / Avg: 576.34 / Max: 690.47 Min: 200.41 / Avg: 651.38 / Max: 755.47 Min: 157.04 / Avg: 656.04 / Max: 755.91 Min: 185.7 / Avg: 663.3 / Max: 752.58 Min: 67.1 / Avg: 671.28 / Max: 820.11
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: Standard EPYC 9684X EPYC 9684X 2P EPYC 9654 EPYC 9654 2P EPYC 9754 EPYC 8534PN EPYC 9754 2P EPYC 8534P Xeon Platinum 8380 Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Max 9468 Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 11 22 33 44 55 SE +/- 0.61, N = 15 SE +/- 0.40, N = 15 SE +/- 0.50, N = 15 SE +/- 0.61, N = 15 SE +/- 0.45, N = 15 SE +/- 0.61, N = 15 SE +/- 0.23, N = 15 SE +/- 0.63, N = 15 SE +/- 0.01, N = 3 SE +/- 0.09, N = 3 SE +/- 0.09, N = 3 SE +/- 0.04, N = 3 SE +/- 0.12, N = 3 SE +/- 0.35, N = 3 SE +/- 0.11, N = 3 SE +/- 0.31, N = 3 SE +/- 0.33, N = 3 SE +/- 0.40, N = 3 SE +/- 0.06, N = 3 SE +/- 0.08, N = 3 SE +/- 0.35, N = 3 SE +/- 0.36, N = 11 21.58 22.79 23.92 25.32 26.45 26.83 27.90 27.93 28.02 28.85 29.02 29.41 29.75 29.76 33.33 33.99 36.00 36.23 36.78 41.08 43.81 47.20 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 Xeon 6766E 2P Xeon 6780E 2P EPYC 9684X EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P 90 180 270 360 450 Min: 16.01 / Avg: 98.3 / Max: 106.86 Min: 18.14 / Avg: 109.41 / Max: 118.96 Min: 43.64 / Avg: 132.23 / Max: 144.22 Min: 43.52 / Avg: 154.49 / Max: 168.64 Min: 19.69 / Avg: 171.56 / Max: 188.27 Min: 26.28 / Avg: 211.37 / Max: 231.46 Min: 53.43 / Avg: 226.19 / Max: 247.32 Min: 104.4 / Avg: 226.53 / Max: 245.57 Min: 106.43 / Avg: 248.46 / Max: 271.28 Min: 31.9 / Avg: 250.92 / Max: 275.39 Min: 40.25 / Avg: 308.06 / Max: 331.38 Min: 54.43 / Avg: 417.34 / Max: 450.65 Min: 103.45 / Avg: 419.57 / Max: 461.32 Min: 60.67 / Avg: 475.62 / Max: 514.58
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: Standard EPYC 9684X EPYC 9684X 2P EPYC 9654 EPYC 9654 2P EPYC 9754 EPYC 8534PN EPYC 8534P EPYC 9754 2P Xeon Platinum 8380 Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Max 9468 Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 11 22 33 44 55 SE +/- 1.36, N = 15 SE +/- 0.76, N = 15 SE +/- 0.91, N = 15 SE +/- 0.92, N = 15 SE +/- 0.63, N = 15 SE +/- 0.82, N = 15 SE +/- 0.87, N = 15 SE +/- 0.27, N = 15 SE +/- 0.01, N = 3 SE +/- 0.11, N = 3 SE +/- 0.10, N = 3 SE +/- 0.05, N = 3 SE +/- 0.13, N = 3 SE +/- 0.40, N = 3 SE +/- 0.09, N = 3 SE +/- 0.27, N = 3 SE +/- 0.26, N = 3 SE +/- 0.31, N = 3 SE +/- 0.05, N = 3 SE +/- 0.05, N = 3 SE +/- 0.18, N = 3 SE +/- 0.16, N = 11 46.88 44.06 42.07 39.80 37.96 37.53 36.08 35.87 35.69 34.66 34.46 34.00 33.61 33.60 30.00 29.43 27.78 27.61 27.18 24.34 22.83 21.20 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: Standard EPYC 9684X EPYC 9684X 2P EPYC 9654 EPYC 9654 2P EPYC 9754 EPYC 8534PN EPYC 9754 2P EPYC 8534P Xeon Platinum 8380 Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Max 9468 Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 11 22 33 44 55 SE +/- 0.61, N = 15 SE +/- 0.40, N = 15 SE +/- 0.50, N = 15 SE +/- 0.61, N = 15 SE +/- 0.45, N = 15 SE +/- 0.61, N = 15 SE +/- 0.23, N = 15 SE +/- 0.63, N = 15 SE +/- 0.01, N = 3 SE +/- 0.09, N = 3 SE +/- 0.09, N = 3 SE +/- 0.04, N = 3 SE +/- 0.12, N = 3 SE +/- 0.35, N = 3 SE +/- 0.11, N = 3 SE +/- 0.31, N = 3 SE +/- 0.33, N = 3 SE +/- 0.40, N = 3 SE +/- 0.06, N = 3 SE +/- 0.08, N = 3 SE +/- 0.35, N = 3 SE +/- 0.36, N = 11 21.58 22.79 23.92 25.32 26.45 26.83 27.90 27.93 28.02 28.85 29.02 29.41 29.75 29.76 33.33 33.99 36.00 36.23 36.78 41.08 43.81 47.20 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OpenFOAM OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Medium Mesh Size - Mesh Time EPYC 9684X Xeon Max 9468 2P EPYC 9684X 2P EPYC 9654 Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8592+ Xeon Max 9480 Xeon Max 9468 Xeon Max 9480 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Platinum 8380 2P EPYC 8534P EPYC 9754 Xeon Platinum 8380 EPYC 8534PN EPYC 9754 2P 40 80 120 160 200 118.54 134.78 135.52 136.40 139.62 140.48 148.40 150.67 153.12 159.31 160.73 162.03 162.32 163.70 167.41 182.24 188.37 192.42 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm
Timed LLVM Compilation This test times how long it takes to compile/build the LLVM compiler stack. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Timed LLVM Compilation 16.0 Build System: Ninja EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X Xeon 6780E 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9654 Xeon 6766E 2P EPYC 9754 Xeon Platinum 8592+ Xeon 6780E Xeon 6766E Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon Max 9480 EPYC 8534P Xeon Max 9468 EPYC 8534PN Xeon Platinum 8380 60 120 180 240 300 SE +/- 0.09, N = 3 SE +/- 0.08, N = 3 SE +/- 0.45, N = 3 SE +/- 0.27, N = 3 SE +/- 0.15, N = 3 SE +/- 0.27, N = 3 SE +/- 0.05, N = 3 SE +/- 0.79, N = 3 SE +/- 0.17, N = 3 SE +/- 0.31, N = 3 SE +/- 1.52, N = 3 SE +/- 0.74, N = 3 SE +/- 0.12, N = 3 SE +/- 0.06, N = 3 SE +/- 0.20, N = 3 SE +/- 0.07, N = 3 SE +/- 0.37, N = 3 SE +/- 0.36, N = 3 SE +/- 0.20, N = 3 SE +/- 0.19, N = 3 SE +/- 0.10, N = 3 SE +/- 1.03, N = 3 89.28 94.81 101.56 106.82 122.31 124.46 125.27 128.07 134.42 135.67 138.57 146.76 165.05 169.45 184.20 187.02 194.38 197.84 205.01 212.48 215.93 298.41 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Timed LLVM Compilation 16.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 Xeon 6766E 2P Xeon 6780E 2P EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 15.37 / Avg: 117.14 / Max: 147.36 Min: 16.99 / Avg: 134 / Max: 170.95 Min: 43.51 / Avg: 148.04 / Max: 239.55 Min: 43.38 / Avg: 172.6 / Max: 293.27 Min: 21.92 / Avg: 221.05 / Max: 334.41 Min: 28.01 / Avg: 241.32 / Max: 332.83 Min: 51.36 / Avg: 243.86 / Max: 291.49 Min: 104.24 / Avg: 248.44 / Max: 496.75 Min: 107.08 / Avg: 281.29 / Max: 590.24 Min: 33.31 / Avg: 289.22 / Max: 411.83 Min: 33.37 / Avg: 291.41 / Max: 377.45 Min: 58.01 / Avg: 298.24 / Max: 377.21 Min: 83.29 / Avg: 311.03 / Max: 377.86 Min: 74.97 / Avg: 314.74 / Max: 377.51 Min: 41.44 / Avg: 382.15 / Max: 678.1 Min: 52.7 / Avg: 427.31 / Max: 682.72 Min: 106.11 / Avg: 438.51 / Max: 581.89 Min: 69.22 / Avg: 491.18 / Max: 825.37 Min: 81.57 / Avg: 519.72 / Max: 754.2 Min: 162.49 / Avg: 544.93 / Max: 754.95 Min: 206.76 / Avg: 582.56 / Max: 754.19 Min: 201.26 / Avg: 591.84 / Max: 754.62
Memcached Memcached is a high performance, distributed memory object caching system. This Memcached test profiles makes use of memtier_benchmark for excuting this CPU/memory-focused server benchmark. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Ops/sec, More Is Better Memcached 1.6.19 Set To Get Ratio: 1:100 EPYC 9684X 2P EPYC 9684X EPYC 9654 2P EPYC 9654 EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 2M 4M 6M 8M 10M SE +/- 27143.58, N = 3 SE +/- 58081.25, N = 3 SE +/- 83562.82, N = 3 SE +/- 29027.89, N = 3 SE +/- 38372.96, N = 3 SE +/- 14591.70, N = 3 SE +/- 36366.89, N = 3 SE +/- 41123.40, N = 3 SE +/- 38410.91, N = 15 SE +/- 101903.23, N = 15 SE +/- 35182.67, N = 3 SE +/- 25335.31, N = 3 SE +/- 21428.06, N = 3 SE +/- 52575.97, N = 3 SE +/- 35300.21, N = 3 SE +/- 24083.78, N = 3 SE +/- 26891.33, N = 3 SE +/- 50660.01, N = 15 SE +/- 19998.04, N = 15 SE +/- 19070.02, N = 15 SE +/- 16926.98, N = 15 SE +/- 16910.46, N = 15 8683410.79 8634258.39 8498306.78 8220895.45 7736018.74 7645325.70 6198256.72 5820618.18 4826284.80 4814941.53 4752673.79 4481667.95 4384040.52 4315467.17 3780240.56 3392842.38 3373323.92 3245908.40 2339570.28 2299150.51 1821144.00 1783869.96 1. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre Ops/sec Per Watt
OpenBenchmarking.org Ops/sec Per Watt, More Is Better Memcached 1.6.19 Set To Get Ratio: 1:100 EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 EPYC 9684X Xeon 6766E Xeon Platinum 8380 EPYC 9754 2P Xeon 6780E Xeon Platinum 8592+ EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P Xeon 6766E 2P Xeon Platinum 8592+ 2P Xeon 6780E 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P 11K 22K 33K 44K 55K 51146.06 44935.74 32910.93 31074.26 25824.73 21456.14 20130.49 19415.88 18800.36 18089.30 17990.14 15985.01 15564.53 15062.13 14272.39 10688.23 8031.40 7668.81 7280.70 6127.33 5975.55 5947.33 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Memcached 1.6.19 CPU Power Consumption Monitor Xeon 6766E EPYC 8534PN Xeon 6780E EPYC 8534P Xeon 6766E 2P EPYC 9754 Xeon Platinum 8380 Xeon 6780E 2P Xeon Platinum 8592+ EPYC 9654 Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9684X EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 110 220 330 440 550 Min: 43.49 / Avg: 107.16 / Max: 120.67 Min: 16.61 / Avg: 113.8 / Max: 137.75 Min: 43.41 / Avg: 124.44 / Max: 142.61 Min: 19.03 / Avg: 137.94 / Max: 159.73 Min: 104.69 / Avg: 222.11 / Max: 257.75 Min: 21.61 / Avg: 235.06 / Max: 274.9 Min: 53.19 / Avg: 239.75 / Max: 291.89 Min: 107.52 / Avg: 250.13 / Max: 283.85 Min: 24.01 / Avg: 262.73 / Max: 308.93 Min: 26.63 / Avg: 264.56 / Max: 307.16 Min: 53.66 / Avg: 287.94 / Max: 334.72 Min: 70.74 / Avg: 291.06 / Max: 331.31 Min: 56.01 / Avg: 302.36 / Max: 343.09 Min: 33.48 / Avg: 334.34 / Max: 390.93 Min: 40.14 / Avg: 393.77 / Max: 460.35 Min: 98.17 / Avg: 450.49 / Max: 528.19 Min: 55.17 / Avg: 472.39 / Max: 548.36 Min: 62.89 / Avg: 492.94 / Max: 577.85 Min: 64.05 / Avg: 543.22 / Max: 637.08 Min: 165.51 / Avg: 545.78 / Max: 623.84 Min: 197.32 / Avg: 550.54 / Max: 613.93 Min: 189.98 / Avg: 567.79 / Max: 640.73
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Standard Xeon Platinum 8380 Xeon Max 9468 Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Max 9480 EPYC 9684X EPYC 9654 Xeon Platinum 8380 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P EPYC 9754 Xeon 6780E EPYC 9684X 2P EPYC 9654 2P Xeon 6766E EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 3 6 9 12 15 SE +/- 0.01476, N = 3 SE +/- 0.02187, N = 3 SE +/- 0.01810, N = 3 SE +/- 0.02572, N = 3 SE +/- 0.03815, N = 5 SE +/- 0.01598, N = 3 SE +/- 0.08822, N = 15 SE +/- 0.01280, N = 3 SE +/- 0.04633, N = 15 SE +/- 0.05392, N = 15 SE +/- 0.03225, N = 3 SE +/- 0.09530, N = 15 SE +/- 0.01358, N = 3 SE +/- 0.06816, N = 3 SE +/- 0.07891, N = 15 SE +/- 0.03249, N = 3 SE +/- 0.14799, N = 15 SE +/- 0.20624, N = 12 SE +/- 0.08614, N = 3 SE +/- 0.10131, N = 15 SE +/- 0.00701, N = 3 SE +/- 0.06840, N = 3 2.98306 3.02897 3.17002 3.18392 3.37034 4.10914 4.46135 4.49317 4.57029 4.74221 5.62517 5.74491 5.83954 6.17796 6.22791 6.97509 7.09682 7.55558 7.68977 8.51808 11.48200 12.43450 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 EPYC 9654 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P 100 200 300 400 500 Min: 15.97 / Avg: 109.73 / Max: 119.37 Min: 17.38 / Avg: 123.79 / Max: 132.65 Min: 43.68 / Avg: 148.15 / Max: 156.57 Min: 43.64 / Avg: 174.72 / Max: 185.33 Min: 15.87 / Avg: 195.33 / Max: 210.52 Min: 54.43 / Avg: 235.28 / Max: 251.36 Min: 29.39 / Avg: 241.91 / Max: 260.94 Min: 31.68 / Avg: 288.98 / Max: 309.48 Min: 104.79 / Avg: 305.05 / Max: 321.54 Min: 108.03 / Avg: 350.61 / Max: 370.78 Min: 39.76 / Avg: 407.81 / Max: 436.29 Min: 101.4 / Avg: 445.3 / Max: 473.57 Min: 55.26 / Avg: 482.11 / Max: 517.72 Min: 60.4 / Avg: 551.5 / Max: 590.7
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Standard Xeon Platinum 8380 Xeon Max 9468 Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Max 9480 EPYC 9684X EPYC 9654 Xeon Platinum 8380 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P EPYC 9754 Xeon 6780E EPYC 9684X 2P EPYC 9654 2P Xeon 6766E EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 70 140 210 280 350 SE +/- 1.65, N = 3 SE +/- 2.40, N = 3 SE +/- 1.80, N = 3 SE +/- 2.52, N = 3 SE +/- 3.31, N = 5 SE +/- 0.95, N = 3 SE +/- 4.35, N = 15 SE +/- 0.64, N = 3 SE +/- 2.24, N = 15 SE +/- 2.40, N = 15 SE +/- 1.01, N = 3 SE +/- 2.72, N = 15 SE +/- 0.40, N = 3 SE +/- 1.78, N = 3 SE +/- 2.12, N = 15 SE +/- 0.67, N = 3 SE +/- 3.21, N = 15 SE +/- 3.84, N = 12 SE +/- 1.47, N = 3 SE +/- 1.44, N = 15 SE +/- 0.05, N = 3 SE +/- 0.44, N = 3 335.15 330.10 315.41 314.06 296.80 243.34 225.31 222.52 219.06 211.19 177.76 174.68 171.23 161.89 160.90 143.36 141.83 133.49 130.06 117.62 87.09 80.42 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Standard Xeon Platinum 8380 Xeon Max 9468 Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Max 9480 EPYC 9684X EPYC 9654 Xeon Platinum 8380 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P EPYC 9754 Xeon 6780E EPYC 9684X 2P EPYC 9654 2P Xeon 6766E EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 3 6 9 12 15 SE +/- 0.01476, N = 3 SE +/- 0.02187, N = 3 SE +/- 0.01810, N = 3 SE +/- 0.02572, N = 3 SE +/- 0.03815, N = 5 SE +/- 0.01598, N = 3 SE +/- 0.08822, N = 15 SE +/- 0.01280, N = 3 SE +/- 0.04633, N = 15 SE +/- 0.05392, N = 15 SE +/- 0.03225, N = 3 SE +/- 0.09530, N = 15 SE +/- 0.01358, N = 3 SE +/- 0.06816, N = 3 SE +/- 0.07891, N = 15 SE +/- 0.03249, N = 3 SE +/- 0.14799, N = 15 SE +/- 0.20624, N = 12 SE +/- 0.08614, N = 3 SE +/- 0.10131, N = 15 SE +/- 0.00701, N = 3 SE +/- 0.06840, N = 3 2.98306 3.02897 3.17002 3.18392 3.37034 4.10914 4.46135 4.49317 4.57029 4.74221 5.62517 5.74491 5.83954 6.17796 6.22791 6.97509 7.09682 7.55558 7.68977 8.51808 11.48200 12.43450 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
NAMD NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ns/day, More Is Better NAMD 3.0b6 Input: ATPase with 327,506 Atoms EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P EPYC 9684X EPYC 9654 EPYC 9754 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8592+ 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8592+ Xeon 6766E Xeon 6780E Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9480 2P Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8490H 2P 5 10 15 20 25 SE +/- 0.08771, N = 3 SE +/- 0.02125, N = 3 SE +/- 0.04751, N = 3 SE +/- 0.02148, N = 6 SE +/- 0.01368, N = 5 SE +/- 0.04917, N = 5 SE +/- 0.26203, N = 15 SE +/- 0.07204, N = 4 SE +/- 0.16058, N = 15 SE +/- 0.14537, N = 15 SE +/- 0.34017, N = 12 SE +/- 0.04338, N = 5 SE +/- 0.00431, N = 4 SE +/- 0.54361, N = 15 SE +/- 0.02593, N = 3 SE +/- 0.17734, N = 15 SE +/- 0.02582, N = 3 SE +/- 0.21567, N = 12 SE +/- 0.44989, N = 15 SE +/- 0.19805, N = 15 SE +/- 0.20408, N = 12 SE +/- 0.17435, N = 15 20.75259 18.97454 17.81105 9.33534 8.18275 7.99756 7.30065 6.72663 5.97130 5.92501 5.85267 5.70672 5.44971 4.64814 3.76811 3.76348 3.60479 2.97021 2.56319 2.55400 2.39770 2.21948 ns/day Per Watt
OpenBenchmarking.org ns/day Per Watt, More Is Better NAMD 3.0b6 Input: ATPase with 327,506 Atoms EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P EPYC 8534PN EPYC 8534P EPYC 9684X EPYC 9754 EPYC 9654 Xeon 6766E Xeon 6780E Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon 6766E 2P Xeon Platinum 8490H Xeon 6780E 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P 0.0189 0.0378 0.0567 0.0756 0.0945 0.084 0.068 0.064 0.056 0.052 0.052 0.048 0.044 0.034 0.029 0.024 0.022 0.020 0.017 0.017 0.014 0.012 0.011 0.011 0.009 0.005 0.004 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better NAMD 3.0b6 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9684X EPYC 9654 Xeon Platinum 8380 EPYC 9754 2P Xeon Platinum 8592+ Xeon 6766E 2P Xeon Max 9468 Xeon Platinum 8490H EPYC 9654 2P Xeon 6780E 2P Xeon Max 9480 EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 16.45 / Avg: 96.77 / Max: 141.98 Min: 17 / Avg: 109.51 / Max: 166.07 Min: 43.66 / Avg: 112.28 / Max: 262.76 Min: 43.48 / Avg: 132.04 / Max: 342.42 Min: 20.68 / Avg: 165.67 / Max: 289.24 Min: 29.28 / Avg: 180.85 / Max: 379.02 Min: 27.01 / Avg: 187.68 / Max: 312.41 Min: 54.68 / Avg: 216.04 / Max: 281.76 Min: 40.88 / Avg: 225.05 / Max: 615.88 Min: 33.34 / Avg: 236.13 / Max: 373.62 Min: 72.63 / Avg: 236.64 / Max: 427.46 Min: 73.54 / Avg: 248.99 / Max: 374.54 Min: 55.01 / Avg: 258.53 / Max: 352.06 Min: 58.84 / Avg: 262.91 / Max: 648.62 Min: 106.87 / Avg: 266.5 / Max: 492.66 Min: 73.99 / Avg: 269.88 / Max: 376.76 Min: 62.44 / Avg: 326.5 / Max: 779.77 Min: 101.59 / Avg: 389.17 / Max: 564.97 Min: 75.47 / Avg: 484.77 / Max: 671.95 Min: 194.24 / Avg: 522.05 / Max: 693.75 Min: 165.98 / Avg: 533.68 / Max: 633.47 Min: 190.7 / Avg: 550.3 / Max: 686.19
Speedb Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Op/s, More Is Better Speedb 2.7 Test: Read While Writing EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 EPYC 9684X EPYC 9654 Xeon 6780E Xeon Platinum 8592+ 2P Xeon 6766E Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 8M 16M 24M 32M 40M SE +/- 803844.71, N = 15 SE +/- 651901.79, N = 15 SE +/- 746932.76, N = 15 SE +/- 223786.94, N = 15 SE +/- 138092.00, N = 3 SE +/- 172052.60, N = 3 SE +/- 118870.12, N = 3 SE +/- 170909.18, N = 3 SE +/- 75416.94, N = 3 SE +/- 195912.00, N = 15 SE +/- 45758.35, N = 3 SE +/- 239363.92, N = 3 SE +/- 197596.27, N = 3 SE +/- 171807.74, N = 3 SE +/- 143570.67, N = 4 SE +/- 88013.69, N = 15 SE +/- 86788.79, N = 3 SE +/- 48351.08, N = 3 SE +/- 62510.72, N = 3 SE +/- 93731.24, N = 7 SE +/- 88257.34, N = 3 SE +/- 48302.92, N = 15 36593204 33307574 30324596 27657776 27630781 22829426 19595777 18024993 17945745 17585737 17548934 17297027 15354400 13859254 12168413 12053901 11882103 11795318 10695360 9988176 8795408 7147120 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread Op/s Per Watt
OpenBenchmarking.org Op/s Per Watt, More Is Better Speedb 2.7 Test: Read While Writing Xeon 6766E EPYC 8534PN EPYC 9754 Xeon 6780E EPYC 9754 2P EPYC 8534P Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 EPYC 9654 2P EPYC 9684X 2P EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Platinum 8380 Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Max 9468 2P 20K 40K 60K 80K 100K 103697.42 97163.65 93506.84 90565.21 85339.03 79851.05 77049.13 69360.79 63472.87 61034.97 57884.56 57032.77 37383.41 32356.04 30648.42 29984.58 27890.69 26509.91 26260.05 23084.61 23023.23 21183.20 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Speedb 2.7 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 17.9 / Avg: 122.29 / Max: 139.21 Min: 17.32 / Avg: 150.95 / Max: 163.43 Min: 43.64 / Avg: 169.23 / Max: 238.92 Min: 43.55 / Avg: 198.15 / Max: 300.83 Min: 22.4 / Avg: 244.15 / Max: 314.31 Min: 52.53 / Avg: 256.25 / Max: 291.99 Min: 26.3 / Avg: 283.98 / Max: 305.25 Min: 34.03 / Avg: 325.5 / Max: 377.13 Min: 52.35 / Avg: 330.55 / Max: 376.19 Min: 78.68 / Avg: 331.78 / Max: 377.28 Min: 74.62 / Avg: 333.11 / Max: 375.39 Min: 32.81 / Avg: 343.59 / Max: 411.18 Min: 107.63 / Avg: 358.61 / Max: 475.78 Min: 108.97 / Avg: 398.75 / Max: 547.56 Min: 41.53 / Avg: 428.8 / Max: 640.33 Min: 56.42 / Avg: 496.84 / Max: 614.4 Min: 105.62 / Avg: 510.96 / Max: 581.73 Min: 78.24 / Avg: 573.79 / Max: 750.97 Min: 64.91 / Avg: 575.41 / Max: 820.95 Min: 210.71 / Avg: 654.26 / Max: 754.28 Min: 164.82 / Avg: 658.68 / Max: 750.75 Min: 210.91 / Avg: 666.91 / Max: 751.13
PyTorch This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org batches/sec, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 64 - Model: ResNet-50 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P EPYC 9654 EPYC 9684X EPYC 8534P EPYC 8534PN Xeon 6780E EPYC 9754 Xeon 6766E EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P 14 28 42 56 70 SE +/- 0.12, N = 3 SE +/- 0.24, N = 3 SE +/- 0.71, N = 3 SE +/- 0.62, N = 3 SE +/- 0.14, N = 3 SE +/- 0.51, N = 3 SE +/- 0.12, N = 3 SE +/- 0.46, N = 4 SE +/- 0.40, N = 15 SE +/- 0.50, N = 3 SE +/- 0.26, N = 3 SE +/- 0.10, N = 3 SE +/- 0.06, N = 3 SE +/- 0.18, N = 3 SE +/- 0.33, N = 5 SE +/- 0.10, N = 3 SE +/- 0.14, N = 3 SE +/- 0.37, N = 12 SE +/- 0.35, N = 12 SE +/- 0.11, N = 3 SE +/- 0.06, N = 3 SE +/- 0.23, N = 9 63.06 56.01 51.55 51.35 48.99 42.35 40.77 38.30 37.75 37.08 35.01 34.71 33.49 33.23 30.69 27.68 26.89 17.28 17.07 16.10 14.84 13.68 batches/sec Per Watt
OpenBenchmarking.org batches/sec Per Watt, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 64 - Model: ResNet-50 EPYC 8534PN EPYC 8534P Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H Xeon 6766E Xeon 6780E Xeon Max 9468 Xeon Max 9480 EPYC 9654 EPYC 9754 EPYC 9684X Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9754 2P EPYC 9654 2P Xeon 6766E 2P Xeon 6780E 2P EPYC 9684X 2P 0.0761 0.1522 0.2283 0.3044 0.3805 0.338 0.308 0.258 0.236 0.223 0.194 0.193 0.193 0.186 0.179 0.175 0.150 0.116 0.080 0.076 0.074 0.066 0.046 0.046 0.045 0.044 0.042 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better PyTorch 2.2.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon Platinum 8380 2P Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 110 220 330 440 550 Min: 15.03 / Avg: 98.23 / Max: 110.34 Min: 19.48 / Avg: 108.57 / Max: 123.57 Min: 43.6 / Avg: 138.69 / Max: 152.97 Min: 21.23 / Avg: 158.01 / Max: 176.03 Min: 43.46 / Avg: 159.15 / Max: 179.26 Min: 28.44 / Avg: 195.3 / Max: 220.5 Min: 53.3 / Avg: 207.69 / Max: 238.28 Min: 35.79 / Avg: 231.06 / Max: 261.25 Min: 35.31 / Avg: 244 / Max: 293.19 Min: 60.41 / Avg: 250.31 / Max: 298.35 Min: 73.38 / Avg: 266.73 / Max: 308.34 Min: 76.06 / Avg: 275.89 / Max: 319.9 Min: 40.2 / Avg: 299.48 / Max: 333.06 Min: 100.89 / Avg: 325.09 / Max: 418.72 Min: 104.56 / Avg: 326.43 / Max: 349.99 Min: 108.31 / Avg: 363.14 / Max: 388.67 Min: 54.65 / Avg: 367.41 / Max: 412.66 Min: 62.55 / Avg: 415.63 / Max: 466.73 Min: 162.87 / Avg: 486.48 / Max: 565.96 Min: 81.92 / Avg: 507.01 / Max: 574.06 Min: 202.91 / Avg: 568.79 / Max: 623.75 Min: 194.23 / Avg: 577.63 / Max: 637.3
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Max 9480 EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 8534PN EPYC 9754 EPYC 8534P EPYC 9654 2P EPYC 9684X 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 40 80 120 160 200 SE +/- 0.19, N = 3 SE +/- 0.06, N = 3 SE +/- 0.21, N = 3 SE +/- 0.32, N = 15 SE +/- 0.09, N = 3 SE +/- 0.33, N = 3 SE +/- 0.03, N = 3 SE +/- 0.21, N = 3 SE +/- 0.28, N = 4 SE +/- 0.39, N = 13 SE +/- 0.36, N = 3 SE +/- 0.18, N = 3 SE +/- 0.31, N = 15 SE +/- 0.40, N = 15 SE +/- 0.07, N = 3 SE +/- 0.56, N = 15 SE +/- 0.98, N = 15 SE +/- 1.27, N = 15 SE +/- 0.35, N = 3 SE +/- 0.34, N = 3 SE +/- 0.28, N = 3 SE +/- 1.15, N = 3 15.04 18.93 19.60 21.22 21.96 23.66 24.11 24.65 25.48 26.09 27.59 29.44 32.23 32.47 32.92 42.51 42.89 55.92 96.69 106.69 167.97 179.14 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 EPYC 9654 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P 110 220 330 440 550 Min: 16.31 / Avg: 120.57 / Max: 131.45 Min: 17.21 / Avg: 132.4 / Max: 145.89 Min: 43.72 / Avg: 151.02 / Max: 162.14 Min: 43.54 / Avg: 183.75 / Max: 198.81 Min: 20.36 / Avg: 200.16 / Max: 217.19 Min: 53.12 / Avg: 250.25 / Max: 269.16 Min: 25.68 / Avg: 250.64 / Max: 270.41 Min: 34.41 / Avg: 294.42 / Max: 320.91 Min: 103.92 / Avg: 301.78 / Max: 322.43 Min: 107.32 / Avg: 350.27 / Max: 375.87 Min: 37.61 / Avg: 403.98 / Max: 434.7 Min: 99.93 / Avg: 462.76 / Max: 500.69 Min: 56.82 / Avg: 486.07 / Max: 528.33 Min: 58.32 / Avg: 553.19 / Max: 601.27
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Max 9480 EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 8534PN EPYC 9754 EPYC 8534P EPYC 9654 2P EPYC 9684X 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 15 30 45 60 75 SE +/- 0.82440, N = 3 SE +/- 0.16471, N = 3 SE +/- 0.54692, N = 3 SE +/- 0.69010, N = 15 SE +/- 0.18826, N = 3 SE +/- 0.58786, N = 3 SE +/- 0.05911, N = 3 SE +/- 0.34209, N = 3 SE +/- 0.42374, N = 4 SE +/- 0.50579, N = 13 SE +/- 0.47575, N = 3 SE +/- 0.20595, N = 3 SE +/- 0.29092, N = 15 SE +/- 0.38216, N = 15 SE +/- 0.06539, N = 3 SE +/- 0.31646, N = 15 SE +/- 0.56636, N = 15 SE +/- 0.40826, N = 15 SE +/- 0.03724, N = 3 SE +/- 0.02972, N = 3 SE +/- 0.00992, N = 3 SE +/- 0.03571, N = 3 66.51620 52.81670 51.01720 47.27030 45.54140 42.27470 41.47450 40.56710 39.25860 38.42400 36.24930 33.96210 31.06440 30.85720 30.37430 23.57860 23.49600 18.01330 10.34250 9.37294 5.95283 5.58241 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Max 9480 EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 8534PN EPYC 9754 EPYC 8534P EPYC 9654 2P EPYC 9684X 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 40 80 120 160 200 SE +/- 0.19, N = 3 SE +/- 0.06, N = 3 SE +/- 0.21, N = 3 SE +/- 0.32, N = 15 SE +/- 0.09, N = 3 SE +/- 0.33, N = 3 SE +/- 0.03, N = 3 SE +/- 0.21, N = 3 SE +/- 0.28, N = 4 SE +/- 0.39, N = 13 SE +/- 0.36, N = 3 SE +/- 0.18, N = 3 SE +/- 0.31, N = 15 SE +/- 0.40, N = 15 SE +/- 0.07, N = 3 SE +/- 0.56, N = 15 SE +/- 0.98, N = 15 SE +/- 1.27, N = 15 SE +/- 0.35, N = 3 SE +/- 0.34, N = 3 SE +/- 0.28, N = 3 SE +/- 1.15, N = 3 15.04 18.93 19.60 21.22 21.96 23.66 24.11 24.65 25.48 26.09 27.59 29.44 32.23 32.47 32.92 42.51 42.89 55.92 96.69 106.69 167.97 179.14 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: bertsquad-12 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P EPYC 9684X EPYC 9654 EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 30 60 90 120 150 SE +/- 0.04, N = 3 SE +/- 0.44, N = 3 SE +/- 0.44, N = 3 SE +/- 1.84, N = 12 SE +/- 0.17, N = 3 SE +/- 0.23, N = 3 SE +/- 0.29, N = 3 SE +/- 1.12, N = 15 SE +/- 0.68, N = 3 SE +/- 0.62, N = 3 SE +/- 0.18, N = 3 SE +/- 0.17, N = 3 SE +/- 0.51, N = 3 SE +/- 1.26, N = 3 SE +/- 0.70, N = 15 SE +/- 2.80, N = 15 SE +/- 2.33, N = 12 SE +/- 0.90, N = 6 SE +/- 0.06, N = 3 SE +/- 0.05, N = 3 SE +/- 6.88, N = 15 SE +/- 6.85, N = 15 39.75 44.60 47.24 48.47 49.64 49.97 54.28 55.90 60.11 60.36 71.56 72.12 80.72 86.65 87.66 90.06 93.64 93.83 111.87 124.06 141.57 155.90 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon 6766E 2P EPYC 9654 Xeon 6780E 2P Xeon Platinum 8380 EPYC 9684X EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P 90 180 270 360 450 Min: 16.97 / Avg: 101.89 / Max: 112.56 Min: 16.8 / Avg: 119.38 / Max: 129.78 Min: 43.62 / Avg: 139.52 / Max: 148.88 Min: 43.51 / Avg: 163.29 / Max: 175 Min: 20.33 / Avg: 188.2 / Max: 202.92 Min: 104.06 / Avg: 210.79 / Max: 227.9 Min: 29.64 / Avg: 222.91 / Max: 240.57 Min: 106.24 / Avg: 233.19 / Max: 250.31 Min: 51.89 / Avg: 246.13 / Max: 266.29 Min: 33.42 / Avg: 261.74 / Max: 283.2 Min: 38.39 / Avg: 302.95 / Max: 323.86 Min: 56.36 / Avg: 405.91 / Max: 434.32 Min: 95.98 / Avg: 424.24 / Max: 455.9 Min: 63.67 / Avg: 460.32 / Max: 497.92
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: bertsquad-12 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P EPYC 9684X EPYC 9654 EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 6 12 18 24 30 SE +/- 0.02370, N = 3 SE +/- 0.21982, N = 3 SE +/- 0.19764, N = 3 SE +/- 0.58074, N = 12 SE +/- 0.07031, N = 3 SE +/- 0.09194, N = 3 SE +/- 0.09869, N = 3 SE +/- 0.35497, N = 15 SE +/- 0.18711, N = 3 SE +/- 0.16743, N = 3 SE +/- 0.03564, N = 3 SE +/- 0.03352, N = 3 SE +/- 0.07825, N = 3 SE +/- 0.16639, N = 3 SE +/- 0.09294, N = 15 SE +/- 0.38561, N = 15 SE +/- 0.26635, N = 12 SE +/- 0.10002, N = 6 SE +/- 0.00509, N = 3 SE +/- 0.00331, N = 3 SE +/- 0.36771, N = 15 SE +/- 0.31004, N = 15 25.15910 22.42510 21.17020 20.87230 20.14530 20.01030 18.42220 17.98730 16.63940 16.57050 13.97450 13.86470 12.38840 11.54470 11.41820 11.27070 10.75188 10.66150 8.93833 8.06013 7.31285 6.60471 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: bertsquad-12 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P EPYC 9684X EPYC 9654 EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 30 60 90 120 150 SE +/- 0.04, N = 3 SE +/- 0.44, N = 3 SE +/- 0.44, N = 3 SE +/- 1.84, N = 12 SE +/- 0.17, N = 3 SE +/- 0.23, N = 3 SE +/- 0.29, N = 3 SE +/- 1.12, N = 15 SE +/- 0.68, N = 3 SE +/- 0.62, N = 3 SE +/- 0.18, N = 3 SE +/- 0.17, N = 3 SE +/- 0.51, N = 3 SE +/- 1.26, N = 3 SE +/- 0.70, N = 15 SE +/- 2.80, N = 15 SE +/- 2.33, N = 12 SE +/- 0.90, N = 6 SE +/- 0.06, N = 3 SE +/- 0.05, N = 3 SE +/- 6.88, N = 15 SE +/- 6.85, N = 15 39.75 44.60 47.24 48.47 49.64 49.97 54.28 55.90 60.11 60.36 71.56 72.12 80.72 86.65 87.66 90.06 93.64 93.83 111.87 124.06 141.57 155.90 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Platinum 8592+ 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Max 9480 2P Xeon 6780E EPYC 8534PN Xeon 6766E EPYC 9684X 2P EPYC 9654 2P EPYC 8534P EPYC 9754 EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 0.8143 1.6286 2.4429 3.2572 4.0715 SE +/- 0.00291, N = 3 SE +/- 0.01594, N = 3 SE +/- 0.00236, N = 3 SE +/- 0.01541, N = 3 SE +/- 0.02719, N = 15 SE +/- 0.01456, N = 3 SE +/- 0.02289, N = 15 SE +/- 0.01405, N = 15 SE +/- 0.01470, N = 3 SE +/- 0.00498, N = 3 SE +/- 0.02493, N = 15 SE +/- 0.01853, N = 5 SE +/- 0.00120, N = 3 SE +/- 0.03117, N = 3 SE +/- 0.00335, N = 3 SE +/- 0.03659, N = 15 SE +/- 0.02209, N = 15 SE +/- 0.02450, N = 4 SE +/- 0.01501, N = 3 SE +/- 0.02347, N = 11 SE +/- 0.03971, N = 3 SE +/- 0.01119, N = 3 1.10042 1.17292 1.17561 1.32454 1.37831 1.41662 1.42849 1.43573 1.48596 1.65270 1.66046 1.71875 2.07288 2.23868 2.24323 2.24645 2.25020 2.25716 2.61653 3.16441 3.41537 3.61920 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P 100 200 300 400 500 Min: 17.14 / Avg: 106.46 / Max: 113.35 Min: 19.26 / Avg: 116.04 / Max: 124.74 Min: 43.72 / Avg: 146.19 / Max: 154.64 Min: 43.58 / Avg: 171.22 / Max: 182.02 Min: 19.01 / Avg: 179.59 / Max: 191.94 Min: 26 / Avg: 222.45 / Max: 240.99 Min: 54.86 / Avg: 232.56 / Max: 248.24 Min: 35.78 / Avg: 264.82 / Max: 285.07 Min: 104.87 / Avg: 303.41 / Max: 320.11 Min: 108.14 / Avg: 349.02 / Max: 367.93 Min: 42.05 / Avg: 397.19 / Max: 423.71 Min: 51.38 / Avg: 431.96 / Max: 458.85 Min: 57.3 / Avg: 472.8 / Max: 505.93 Min: 65.24 / Avg: 542.12 / Max: 580.41
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Platinum 8592+ 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Max 9480 2P Xeon 6780E EPYC 9684X 2P EPYC 8534PN Xeon 6766E EPYC 9654 2P EPYC 8534P EPYC 9754 EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 200 400 600 800 1000 SE +/- 2.43, N = 3 SE +/- 11.46, N = 3 SE +/- 1.66, N = 3 SE +/- 8.88, N = 3 SE +/- 12.91, N = 15 SE +/- 7.34, N = 3 SE +/- 10.46, N = 15 SE +/- 6.50, N = 15 SE +/- 6.60, N = 3 SE +/- 1.83, N = 3 SE +/- 9.50, N = 15 SE +/- 6.21, N = 5 SE +/- 0.28, N = 3 SE +/- 7.76, N = 15 SE +/- 6.13, N = 3 SE +/- 0.67, N = 3 SE +/- 4.37, N = 15 SE +/- 4.84, N = 4 SE +/- 2.19, N = 3 SE +/- 2.36, N = 11 SE +/- 3.37, N = 3 SE +/- 0.86, N = 3 908.19 852.33 849.56 754.81 728.47 705.61 701.72 696.98 672.92 604.76 604.05 581.77 482.21 446.81 446.69 445.62 444.85 443.06 382.10 316.11 292.79 276.24 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Platinum 8592+ 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Max 9480 2P Xeon 6780E EPYC 8534PN Xeon 6766E EPYC 9684X 2P EPYC 9654 2P EPYC 8534P EPYC 9754 EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 0.8143 1.6286 2.4429 3.2572 4.0715 SE +/- 0.00291, N = 3 SE +/- 0.01594, N = 3 SE +/- 0.00236, N = 3 SE +/- 0.01541, N = 3 SE +/- 0.02719, N = 15 SE +/- 0.01456, N = 3 SE +/- 0.02289, N = 15 SE +/- 0.01405, N = 15 SE +/- 0.01470, N = 3 SE +/- 0.00498, N = 3 SE +/- 0.02493, N = 15 SE +/- 0.01853, N = 5 SE +/- 0.00120, N = 3 SE +/- 0.03117, N = 3 SE +/- 0.00335, N = 3 SE +/- 0.03659, N = 15 SE +/- 0.02209, N = 15 SE +/- 0.02450, N = 4 SE +/- 0.01501, N = 3 SE +/- 0.02347, N = 11 SE +/- 0.03971, N = 3 SE +/- 0.01119, N = 3 1.10042 1.17292 1.17561 1.32454 1.37831 1.41662 1.42849 1.43573 1.48596 1.65270 1.66046 1.71875 2.07288 2.23868 2.24323 2.24645 2.25020 2.25716 2.61653 3.16441 3.41537 3.61920 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
Llamafile Mozilla's Llamafile allows distributing and running large language models (LLMs) as a single file. Llamafile aims to make open-source LLMs more accessible to developers and users. Llamafile supports a variety of models, CPUs and GPUs, and other options. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.7 Test: wizardcoder-python-34b-v1.0.Q6_K - Acceleration: CPU EPYC 9684X EPYC 9654 EPYC 9754 Xeon Platinum 8490H Xeon 6780E Xeon 6766E EPYC 9654 2P EPYC 9684X 2P EPYC 8534P EPYC 8534PN EPYC 9754 2P Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Max 9480 Xeon Max 9468 2 4 6 8 10 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.12, N = 6 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 8.48 8.35 7.47 5.85 5.58 5.39 5.24 5.24 5.16 5.11 4.63 4.41 4.08 3.82 3.10 3.05 2.67 2.56 1.77 1.54 1.54 1.47 Tokens Per Second Per Watt
OpenBenchmarking.org Tokens Per Second Per Watt, More Is Better Llamafile 0.7 Test: wizardcoder-python-34b-v1.0.Q6_K - Acceleration: CPU EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon 6766E EPYC 9684X Xeon 6780E Xeon Platinum 8380 Xeon Platinum 8490H Xeon Platinum 8592+ EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 0.0099 0.0198 0.0297 0.0396 0.0495 0.044 0.038 0.037 0.034 0.031 0.029 0.028 0.018 0.018 0.013 0.011 0.011 0.009 0.007 0.007 0.006 0.006 0.005 0.005 0.005 0.003 0.002 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Llamafile 0.7 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Max 9468 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 16.37 / Avg: 115.77 / Max: 139.52 Min: 17.14 / Avg: 135.55 / Max: 164.93 Min: 43.73 / Avg: 175.78 / Max: 254.24 Min: 43.53 / Avg: 201.04 / Max: 326.57 Min: 21.63 / Avg: 203.84 / Max: 298.7 Min: 30.07 / Avg: 245.28 / Max: 317.66 Min: 51.16 / Avg: 250.18 / Max: 288.75 Min: 29.28 / Avg: 296.68 / Max: 399.75 Min: 72.56 / Avg: 306.79 / Max: 356.97 Min: 33.89 / Avg: 307.46 / Max: 338.87 Min: 53.92 / Avg: 317.1 / Max: 376.78 Min: 86.22 / Avg: 325.29 / Max: 362.03 Min: 105.02 / Avg: 350.49 / Max: 483.96 Min: 108.26 / Avg: 383.74 / Max: 590.61 Min: 41.26 / Avg: 412.48 / Max: 564 Min: 58.45 / Avg: 480.08 / Max: 600.8 Min: 101.74 / Avg: 484.21 / Max: 566.46 Min: 65.77 / Avg: 568.13 / Max: 708.34 Min: 85.64 / Avg: 568.91 / Max: 665.68 Min: 170.34 / Avg: 630.52 / Max: 723.74 Min: 203.4 / Avg: 633.28 / Max: 704.68 Min: 184.95 / Avg: 639.77 / Max: 696.32
Timed Godot Game Engine Compilation This test times how long it takes to compile the Godot Game Engine. Godot is a popular, open-source, cross-platform 2D/3D game engine and is built using the SCons build system and targeting the X11 platform. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Timed Godot Game Engine Compilation 4.0 Time To Compile EPYC 9684X 2P EPYC 9654 2P EPYC 9684X Xeon Platinum 8592+ 2P EPYC 9654 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Platinum 8592+ Xeon Max 9468 2P EPYC 9754 2P EPYC 9754 Xeon Platinum 8490H Xeon Max 9480 Xeon 6780E 2P Xeon Max 9468 Xeon Platinum 8380 2P Xeon 6780E EPYC 8534P EPYC 8534PN Xeon 6766E 2P Xeon 6766E Xeon Platinum 8380 40 80 120 160 200 SE +/- 0.62, N = 3 SE +/- 0.06, N = 3 SE +/- 0.12, N = 3 SE +/- 0.29, N = 3 SE +/- 0.16, N = 3 SE +/- 0.56, N = 3 SE +/- 1.69, N = 3 SE +/- 0.91, N = 3 SE +/- 0.56, N = 3 SE +/- 0.55, N = 3 SE +/- 0.20, N = 3 SE +/- 0.76, N = 3 SE +/- 0.32, N = 3 SE +/- 1.40, N = 3 SE +/- 0.18, N = 3 SE +/- 0.69, N = 3 SE +/- 0.73, N = 3 SE +/- 0.32, N = 3 SE +/- 0.16, N = 3 SE +/- 0.64, N = 3 SE +/- 0.16, N = 3 SE +/- 0.22, N = 3 96.16 102.58 103.35 103.38 111.69 118.57 122.70 122.72 123.88 126.11 128.28 140.23 143.81 148.99 149.38 151.54 152.23 154.01 159.62 160.46 165.41 187.88 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Timed Godot Game Engine Compilation 4.0 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 Xeon 6766E 2P EPYC 9654 Xeon 6780E 2P Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9480 Xeon Max 9468 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 16.76 / Avg: 97.75 / Max: 142.01 Min: 43.52 / Avg: 109.04 / Max: 243.23 Min: 18.58 / Avg: 110.14 / Max: 169.2 Min: 43.34 / Avg: 124.67 / Max: 295.77 Min: 18.92 / Avg: 152.32 / Max: 333.75 Min: 103.92 / Avg: 184.93 / Max: 507.62 Min: 24.79 / Avg: 188.63 / Max: 336.96 Min: 106.86 / Avg: 202.24 / Max: 626.79 Min: 50.66 / Avg: 217.11 / Max: 290.99 Min: 33.53 / Avg: 222.04 / Max: 410.61 Min: 33.4 / Avg: 236.2 / Max: 374.96 Min: 55.69 / Avg: 255.97 / Max: 375.31 Min: 39.69 / Avg: 276.98 / Max: 673.64 Min: 82.82 / Avg: 282.51 / Max: 375.66 Min: 76.87 / Avg: 285.63 / Max: 375.65 Min: 58.53 / Avg: 338.4 / Max: 689.21 Min: 97.52 / Avg: 352.55 / Max: 559.9 Min: 65.36 / Avg: 383.39 / Max: 820.55 Min: 82.98 / Avg: 430.41 / Max: 750.75 Min: 154.59 / Avg: 461.39 / Max: 752.04 Min: 197.93 / Avg: 519.6 / Max: 744.47 Min: 199.22 / Avg: 527.95 / Max: 724.51
PyTorch This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org batches/sec, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 256 - Model: ResNet-50 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 2P EPYC 9654 EPYC 9684X EPYC 8534P EPYC 8534PN Xeon 6780E EPYC 9754 Xeon 6766E EPYC 9654 2P EPYC 9684X 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P 14 28 42 56 70 SE +/- 0.64, N = 3 SE +/- 0.52, N = 3 SE +/- 0.76, N = 3 SE +/- 0.62, N = 3 SE +/- 0.04, N = 3 SE +/- 0.52, N = 3 SE +/- 0.29, N = 3 SE +/- 0.45, N = 3 SE +/- 0.49, N = 3 SE +/- 0.23, N = 3 SE +/- 0.21, N = 3 SE +/- 0.09, N = 3 SE +/- 0.18, N = 3 SE +/- 0.13, N = 3 SE +/- 0.13, N = 3 SE +/- 0.08, N = 3 SE +/- 0.12, N = 3 SE +/- 0.37, N = 12 SE +/- 0.32, N = 12 SE +/- 0.12, N = 3 SE +/- 0.11, N = 3 SE +/- 0.28, N = 12 63.60 54.84 52.97 51.56 48.89 41.68 41.45 37.91 37.44 36.97 34.90 34.90 33.68 33.30 30.86 27.78 26.82 17.41 16.97 16.22 14.67 13.11 batches/sec Per Watt
OpenBenchmarking.org batches/sec Per Watt, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 256 - Model: ResNet-50 EPYC 8534PN EPYC 8534P Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9468 Xeon 6780E Xeon 6766E Xeon Max 9480 EPYC 9654 EPYC 9754 EPYC 9684X Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P EPYC 9684X 2P 0.0752 0.1504 0.2256 0.3008 0.376 0.334 0.305 0.293 0.235 0.218 0.197 0.194 0.194 0.189 0.179 0.176 0.150 0.113 0.082 0.077 0.073 0.066 0.048 0.045 0.045 0.044 0.040 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better PyTorch 2.2.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 110 220 330 440 550 Min: 17.28 / Avg: 99.64 / Max: 110.85 Min: 19.91 / Avg: 110.56 / Max: 125.45 Min: 43.68 / Avg: 138.47 / Max: 152.75 Min: 21.47 / Avg: 157.48 / Max: 175.9 Min: 43.38 / Avg: 159.18 / Max: 178.76 Min: 30.18 / Avg: 194.95 / Max: 219.41 Min: 55.01 / Avg: 207.81 / Max: 238.76 Min: 18.63 / Avg: 215.65 / Max: 294.3 Min: 32.73 / Avg: 232.73 / Max: 262.5 Min: 61.04 / Avg: 251.84 / Max: 296.62 Min: 75.38 / Avg: 268.45 / Max: 310.8 Min: 89.32 / Avg: 273.3 / Max: 321.11 Min: 40.16 / Avg: 300.12 / Max: 342.36 Min: 105.49 / Avg: 325.94 / Max: 350.34 Min: 101.92 / Avg: 326.58 / Max: 422.87 Min: 108.2 / Avg: 363.82 / Max: 389.75 Min: 55.56 / Avg: 365.56 / Max: 407.37 Min: 61.35 / Avg: 419.25 / Max: 473.7 Min: 165.22 / Avg: 486.52 / Max: 567.17 Min: 72.64 / Avg: 507.69 / Max: 575.72 Min: 201.49 / Avg: 569.91 / Max: 625.34 Min: 193.43 / Avg: 577.93 / Max: 636.86
OpenRadioss OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: INIVOL and Fluid Structure Interaction Drop Container EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P EPYC 9684X Xeon Platinum 8592+ 2P EPYC 9654 EPYC 9754 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 40 80 120 160 200 SE +/- 0.43, N = 3 SE +/- 0.14, N = 3 SE +/- 0.12, N = 3 SE +/- 0.93, N = 3 SE +/- 0.30, N = 3 SE +/- 0.57, N = 3 SE +/- 0.46, N = 3 SE +/- 0.17, N = 3 SE +/- 0.09, N = 3 SE +/- 0.07, N = 3 SE +/- 0.12, N = 3 SE +/- 0.50, N = 3 SE +/- 0.19, N = 3 SE +/- 0.06, N = 3 SE +/- 0.08, N = 3 SE +/- 0.37, N = 3 SE +/- 0.63, N = 3 SE +/- 0.23, N = 3 79.42 85.89 90.85 94.76 100.00 100.24 106.54 107.36 110.56 114.29 128.74 135.56 140.43 146.03 160.56 176.87 187.03 204.21 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenRadioss 2023.09.15 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 18.08 / Avg: 129.32 / Max: 137.09 Min: 18.75 / Avg: 147.36 / Max: 159.1 Min: 52.13 / Avg: 259.44 / Max: 290.81 Min: 21.72 / Avg: 297.41 / Max: 340.37 Min: 27.14 / Avg: 303.45 / Max: 345.2 Min: 23.22 / Avg: 328.15 / Max: 377.07 Min: 60.99 / Avg: 331.72 / Max: 375.56 Min: 75.87 / Avg: 334.64 / Max: 374.39 Min: 72.45 / Avg: 335.98 / Max: 368.12 Min: 34.68 / Avg: 370.43 / Max: 409.62 Min: 99.13 / Avg: 505.66 / Max: 581.63 Min: 41.48 / Avg: 554.53 / Max: 680.58 Min: 57.91 / Avg: 595.98 / Max: 703.76 Min: 67.07 / Avg: 633.7 / Max: 776.53 Min: 175.02 / Avg: 647.82 / Max: 749.74 Min: 215.12 / Avg: 658.11 / Max: 751.86 Min: 204.9 / Avg: 661.11 / Max: 752.72 Min: 65.3 / Avg: 702.37 / Max: 819.28
QMCPACK QMCPACK is a modern high-performance open-source Quantum Monte Carlo (QMC) simulation code making use of MPI for this benchmark of the H20 example code. QMCPACK is an open-source production level many-body ab initio Quantum Monte Carlo code for computing the electronic structure of atoms, molecules, and solids. QMCPACK is supported by the U.S. Department of Energy. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Total Execution Time - Seconds, Fewer Is Better QMCPACK 3.17.1 Input: Li2_STO_ae EPYC 9684X 2P EPYC 9684X Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P EPYC 9654 2P Xeon Platinum 8592+ Xeon Max 9480 2P EPYC 9654 Xeon Platinum 8490H EPYC 9754 EPYC 9754 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8380 40 80 120 160 200 SE +/- 0.06, N = 3 SE +/- 1.06, N = 3 SE +/- 0.93, N = 3 SE +/- 1.11, N = 3 SE +/- 0.88, N = 3 SE +/- 0.23, N = 3 SE +/- 1.03, N = 3 SE +/- 1.11, N = 5 SE +/- 0.54, N = 3 SE +/- 0.89, N = 3 SE +/- 0.63, N = 3 SE +/- 0.42, N = 3 SE +/- 1.04, N = 3 SE +/- 1.20, N = 3 SE +/- 0.50, N = 3 SE +/- 1.13, N = 3 SE +/- 0.40, N = 3 SE +/- 1.47, N = 7 91.97 97.00 97.91 101.63 102.12 102.98 105.98 106.27 108.02 109.30 110.80 111.15 117.81 121.19 131.37 137.23 151.27 163.75 1. (CXX) g++ options: -fopenmp -foffload=disable -finline-limit=1000 -fstrict-aliasing -funroll-all-loops -ffast-math -march=native -O3 -lm -ldl CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better QMCPACK 3.17.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon Platinum 8380 EPYC 9654 EPYC 9754 Xeon Max 9468 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 EPYC 9684X Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.04 / Avg: 132.24 / Max: 137.56 Min: 17.96 / Avg: 154.43 / Max: 159.92 Min: 53.73 / Avg: 263.86 / Max: 290.45 Min: 28.38 / Avg: 316.1 / Max: 334.43 Min: 21.99 / Avg: 319.33 / Max: 338.99 Min: 75.52 / Avg: 323.64 / Max: 364.79 Min: 23.78 / Avg: 335.13 / Max: 377.2 Min: 59.34 / Avg: 337.54 / Max: 376.68 Min: 80.16 / Avg: 338.77 / Max: 371.37 Min: 33.88 / Avg: 391.36 / Max: 409.49 Min: 104.4 / Avg: 524.73 / Max: 581.03 Min: 44.37 / Avg: 628.81 / Max: 690.2 Min: 54.02 / Avg: 635.44 / Max: 688.42 Min: 63.82 / Avg: 660.86 / Max: 753.6 Min: 169.74 / Avg: 670.52 / Max: 753.5 Min: 208.9 / Avg: 673.9 / Max: 748.46 Min: 206.85 / Avg: 676.4 / Max: 753.87 Min: 66.83 / Avg: 767.28 / Max: 818.97
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: fcn-resnet101-11 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ EPYC 9754 Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 EPYC 8534P EPYC 9754 2P EPYC 8534PN Xeon Max 9468 2P Xeon Max 9480 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P EPYC 9654 EPYC 9684X EPYC 9654 2P EPYC 9684X 2P 300 600 900 1200 1500 SE +/- 0.74, N = 3 SE +/- 1.91, N = 3 SE +/- 4.07, N = 3 SE +/- 0.94, N = 3 SE +/- 2.36, N = 15 SE +/- 0.96, N = 3 SE +/- 2.55, N = 3 SE +/- 1.73, N = 3 SE +/- 2.21, N = 3 SE +/- 3.54, N = 3 SE +/- 5.48, N = 3 SE +/- 3.59, N = 15 SE +/- 5.57, N = 13 SE +/- 6.06, N = 3 SE +/- 0.63, N = 3 SE +/- 0.61, N = 3 SE +/- 3.30, N = 3 SE +/- 6.08, N = 15 SE +/- 9.04, N = 12 SE +/- 13.72, N = 15 SE +/- 19.12, N = 12 SE +/- 1.59, N = 3 302.27 319.35 327.88 329.04 345.94 352.74 363.35 380.90 382.11 388.74 397.43 443.96 460.67 475.25 515.08 562.40 587.00 608.09 800.23 801.96 1094.26 1165.93 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P 100 200 300 400 500 Min: 15.82 / Avg: 105.18 / Max: 117.88 Min: 19.36 / Avg: 129.39 / Max: 140.24 Min: 43.69 / Avg: 159.26 / Max: 172.23 Min: 43.63 / Avg: 190.5 / Max: 207.06 Min: 22.2 / Avg: 210.78 / Max: 229.48 Min: 26.58 / Avg: 226.63 / Max: 248.05 Min: 53.76 / Avg: 239.05 / Max: 256.78 Min: 33.01 / Avg: 268.04 / Max: 290.06 Min: 107.09 / Avg: 366.41 / Max: 399.46 Min: 109.34 / Avg: 411.87 / Max: 447.93 Min: 40.33 / Avg: 426.23 / Max: 461.75 Min: 99.33 / Avg: 452.18 / Max: 484 Min: 57.08 / Avg: 475.82 / Max: 514.33 Min: 66.22 / Avg: 538.69 / Max: 582.15
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: fcn-resnet101-11 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ EPYC 9754 Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 EPYC 8534P EPYC 9754 2P EPYC 8534PN Xeon Max 9468 2P Xeon Max 9480 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P EPYC 9684X EPYC 9654 EPYC 9654 2P EPYC 9684X 2P 0.7444 1.4888 2.2332 2.9776 3.722 SE +/- 0.008053, N = 3 SE +/- 0.018814, N = 3 SE +/- 0.038257, N = 3 SE +/- 0.008707, N = 3 SE +/- 0.019617, N = 15 SE +/- 0.007683, N = 3 SE +/- 0.019473, N = 3 SE +/- 0.011894, N = 3 SE +/- 0.015170, N = 3 SE +/- 0.023203, N = 3 SE +/- 0.034519, N = 3 SE +/- 0.018603, N = 15 SE +/- 0.027020, N = 13 SE +/- 0.026516, N = 3 SE +/- 0.002384, N = 3 SE +/- 0.001935, N = 3 SE +/- 0.009611, N = 3 SE +/- 0.016386, N = 15 SE +/- 0.020352, N = 15 SE +/- 0.013875, N = 12 SE +/- 0.016597, N = 12 SE +/- 0.001170, N = 3 3.308360 3.131560 3.050850 3.039220 2.892530 2.834980 2.752420 2.625440 2.617210 2.572780 2.517090 2.254530 2.174630 2.104800 1.941440 1.778090 1.703650 1.646740 1.251790 1.251350 0.917043 0.857685 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: fcn-resnet101-11 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ EPYC 9754 Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 EPYC 8534P EPYC 9754 2P EPYC 8534PN Xeon Max 9468 2P Xeon Max 9480 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P EPYC 9654 EPYC 9684X EPYC 9654 2P EPYC 9684X 2P 300 600 900 1200 1500 SE +/- 0.74, N = 3 SE +/- 1.91, N = 3 SE +/- 4.07, N = 3 SE +/- 0.94, N = 3 SE +/- 2.36, N = 15 SE +/- 0.96, N = 3 SE +/- 2.55, N = 3 SE +/- 1.73, N = 3 SE +/- 2.21, N = 3 SE +/- 3.54, N = 3 SE +/- 5.48, N = 3 SE +/- 3.59, N = 15 SE +/- 5.57, N = 13 SE +/- 6.06, N = 3 SE +/- 0.63, N = 3 SE +/- 0.61, N = 3 SE +/- 3.30, N = 3 SE +/- 6.08, N = 15 SE +/- 9.04, N = 12 SE +/- 13.72, N = 15 SE +/- 19.12, N = 12 SE +/- 1.59, N = 3 302.27 319.35 327.88 329.04 345.94 352.74 363.35 380.90 382.11 388.74 397.43 443.96 460.67 475.25 515.08 562.40 587.00 608.09 800.23 801.96 1094.26 1165.93 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: CaffeNet 12-int8 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 Xeon Platinum 8490H 2P EPYC 8534PN EPYC 8534P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9754 EPYC 9654 2P EPYC 9684X 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 3 6 9 12 15 SE +/- 0.01491, N = 4 SE +/- 0.00936, N = 3 SE +/- 0.00184, N = 3 SE +/- 0.00884, N = 3 SE +/- 0.01053, N = 3 SE +/- 0.00943, N = 3 SE +/- 0.00701, N = 3 SE +/- 0.01586, N = 15 SE +/- 0.01428, N = 15 SE +/- 0.01492, N = 3 SE +/- 0.01684, N = 6 SE +/- 0.00301, N = 3 SE +/- 0.00720, N = 3 SE +/- 0.00871, N = 3 SE +/- 0.01454, N = 15 SE +/- 0.02296, N = 15 SE +/- 0.04200, N = 3 SE +/- 0.03210, N = 15 SE +/- 0.04327, N = 3 SE +/- 0.04869, N = 15 SE +/- 0.11179, N = 3 SE +/- 0.07940, N = 3 1.20298 1.27368 1.30173 1.45280 1.47249 1.55821 1.64919 1.69307 1.70579 1.76353 1.80573 1.82443 2.02659 2.08780 2.21020 3.17064 3.21781 3.69602 5.24051 5.66426 12.12680 12.30210 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P 100 200 300 400 500 Min: 17.47 / Avg: 111.54 / Max: 119.32 Min: 16.81 / Avg: 122.1 / Max: 130.96 Min: 43.72 / Avg: 141.66 / Max: 149.72 Min: 43.54 / Avg: 166.24 / Max: 176.01 Min: 19.98 / Avg: 184.69 / Max: 197.9 Min: 25.06 / Avg: 230.17 / Max: 246.35 Min: 53.65 / Avg: 232.18 / Max: 246.53 Min: 35.49 / Avg: 275.71 / Max: 296 Min: 104.72 / Avg: 302.14 / Max: 319.15 Min: 108.06 / Avg: 346.02 / Max: 366.29 Min: 39.85 / Avg: 399.31 / Max: 425.54 Min: 98.39 / Avg: 431.89 / Max: 460.51 Min: 57.19 / Avg: 475.21 / Max: 509.74 Min: 66.9 / Avg: 542.3 / Max: 581.73
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: CaffeNet 12-int8 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 Xeon Platinum 8490H 2P EPYC 8534PN EPYC 8534P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9754 EPYC 9654 2P EPYC 9684X 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 200 400 600 800 1000 SE +/- 10.06, N = 4 SE +/- 5.72, N = 3 SE +/- 1.09, N = 3 SE +/- 4.16, N = 3 SE +/- 4.87, N = 3 SE +/- 3.90, N = 3 SE +/- 2.57, N = 3 SE +/- 5.53, N = 15 SE +/- 4.99, N = 15 SE +/- 4.81, N = 3 SE +/- 5.34, N = 6 SE +/- 0.91, N = 3 SE +/- 1.77, N = 3 SE +/- 1.98, N = 3 SE +/- 3.02, N = 15 SE +/- 2.35, N = 15 SE +/- 4.10, N = 3 SE +/- 2.41, N = 15 SE +/- 1.57, N = 3 SE +/- 1.50, N = 15 SE +/- 0.77, N = 3 SE +/- 0.53, N = 3 830.73 784.30 767.48 687.67 678.48 641.00 605.59 590.71 586.18 566.38 553.47 547.52 492.89 478.47 452.32 315.36 310.59 270.63 190.78 176.64 82.44 81.26 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: CaffeNet 12-int8 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 Xeon Platinum 8490H 2P EPYC 8534PN EPYC 8534P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9754 EPYC 9654 2P EPYC 9684X 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 3 6 9 12 15 SE +/- 0.01491, N = 4 SE +/- 0.00936, N = 3 SE +/- 0.00184, N = 3 SE +/- 0.00884, N = 3 SE +/- 0.01053, N = 3 SE +/- 0.00943, N = 3 SE +/- 0.00701, N = 3 SE +/- 0.01586, N = 15 SE +/- 0.01428, N = 15 SE +/- 0.01492, N = 3 SE +/- 0.01684, N = 6 SE +/- 0.00301, N = 3 SE +/- 0.00720, N = 3 SE +/- 0.00871, N = 3 SE +/- 0.01454, N = 15 SE +/- 0.02296, N = 15 SE +/- 0.04200, N = 3 SE +/- 0.03210, N = 15 SE +/- 0.04327, N = 3 SE +/- 0.04869, N = 15 SE +/- 0.11179, N = 3 SE +/- 0.07940, N = 3 1.20298 1.27368 1.30173 1.45280 1.47249 1.55821 1.64919 1.69307 1.70579 1.76353 1.80573 1.82443 2.02659 2.08780 2.21020 3.17064 3.21781 3.69602 5.24051 5.66426 12.12680 12.30210 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: T5 Encoder - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P EPYC 9684X EPYC 9684X 2P Xeon Max 9468 Xeon 6780E Xeon Max 9480 Xeon 6780E 2P Xeon 6766E Xeon Max 9468 2P EPYC 9654 Xeon Max 9480 2P EPYC 9654 2P Xeon 6766E 2P EPYC 9754 EPYC 9754 2P EPYC 8534PN EPYC 8534P 2 4 6 8 10 SE +/- 0.00472, N = 3 SE +/- 0.00659, N = 3 SE +/- 0.02371, N = 3 SE +/- 0.03065, N = 15 SE +/- 0.06148, N = 15 SE +/- 0.00316, N = 3 SE +/- 0.01718, N = 3 SE +/- 0.06302, N = 15 SE +/- 0.08827, N = 15 SE +/- 0.00392, N = 3 SE +/- 0.01103, N = 3 SE +/- 0.02282, N = 3 SE +/- 0.00782, N = 3 SE +/- 0.03283, N = 3 SE +/- 0.03084, N = 3 SE +/- 0.02608, N = 3 SE +/- 0.06624, N = 3 SE +/- 0.56391, N = 15 SE +/- 0.06745, N = 3 SE +/- 0.07124, N = 3 SE +/- 0.03354, N = 3 SE +/- 0.06404, N = 13 2.16966 2.79694 2.80263 3.34966 3.43125 3.46571 3.57543 3.80792 4.57581 4.59703 4.64757 4.74859 4.80725 5.02267 5.34656 5.39123 5.98436 6.14327 7.63337 7.84065 7.86015 7.87325 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8380 EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P 60 120 180 240 300 Min: 15.64 / Avg: 68.55 / Max: 73.08 Min: 16.9 / Avg: 70.5 / Max: 76.74 Min: 43.7 / Avg: 85.81 / Max: 89.61 Min: 43.39 / Avg: 88.68 / Max: 93.08 Min: 20.82 / Avg: 94.1 / Max: 99.49 Min: 26.07 / Avg: 125.08 / Max: 132.3 Min: 34.63 / Avg: 146.68 / Max: 155.53 Min: 105.79 / Avg: 152.81 / Max: 159.65 Min: 75.92 / Avg: 152.85 / Max: 173.61 Min: 52.27 / Avg: 202.48 / Max: 217.49 Min: 40.98 / Avg: 208.45 / Max: 223.35 Min: 58.06 / Avg: 256.93 / Max: 271.34 Min: 34.94 / Avg: 287.83 / Max: 306.6 Min: 99.68 / Avg: 289.97 / Max: 319.08
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: T5 Encoder - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P EPYC 9684X EPYC 9684X 2P Xeon Max 9468 Xeon 6780E Xeon Max 9480 Xeon 6780E 2P Xeon 6766E Xeon Max 9468 2P EPYC 9654 Xeon Max 9480 2P Xeon 6766E 2P EPYC 9654 2P EPYC 9754 EPYC 9754 2P EPYC 8534PN EPYC 8534P 100 200 300 400 500 SE +/- 1.00, N = 3 SE +/- 0.84, N = 3 SE +/- 3.01, N = 3 SE +/- 2.62, N = 15 SE +/- 4.93, N = 15 SE +/- 0.26, N = 3 SE +/- 1.35, N = 3 SE +/- 4.17, N = 15 SE +/- 3.95, N = 15 SE +/- 0.17, N = 3 SE +/- 0.51, N = 3 SE +/- 1.00, N = 3 SE +/- 0.35, N = 3 SE +/- 1.31, N = 3 SE +/- 1.08, N = 3 SE +/- 0.89, N = 3 SE +/- 11.98, N = 15 SE +/- 1.86, N = 3 SE +/- 1.16, N = 3 SE +/- 1.15, N = 3 SE +/- 0.54, N = 3 SE +/- 0.98, N = 13 460.78 357.36 356.76 298.67 292.52 288.47 279.64 263.52 219.53 217.45 215.13 210.37 207.94 199.08 187.01 185.46 178.11 167.11 131.00 127.54 127.21 127.09 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: T5 Encoder - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P EPYC 9684X EPYC 9684X 2P Xeon Max 9468 Xeon 6780E Xeon Max 9480 Xeon 6780E 2P Xeon 6766E Xeon Max 9468 2P EPYC 9654 Xeon Max 9480 2P EPYC 9654 2P Xeon 6766E 2P EPYC 9754 EPYC 9754 2P EPYC 8534PN EPYC 8534P 2 4 6 8 10 SE +/- 0.00472, N = 3 SE +/- 0.00659, N = 3 SE +/- 0.02371, N = 3 SE +/- 0.03065, N = 15 SE +/- 0.06148, N = 15 SE +/- 0.00316, N = 3 SE +/- 0.01718, N = 3 SE +/- 0.06302, N = 15 SE +/- 0.08827, N = 15 SE +/- 0.00392, N = 3 SE +/- 0.01103, N = 3 SE +/- 0.02282, N = 3 SE +/- 0.00782, N = 3 SE +/- 0.03283, N = 3 SE +/- 0.03084, N = 3 SE +/- 0.02608, N = 3 SE +/- 0.06624, N = 3 SE +/- 0.56391, N = 15 SE +/- 0.06745, N = 3 SE +/- 0.07124, N = 3 SE +/- 0.03354, N = 3 SE +/- 0.06404, N = 13 2.16966 2.79694 2.80263 3.34966 3.43125 3.46571 3.57543 3.80792 4.57581 4.59703 4.64757 4.74859 4.80725 5.02267 5.34656 5.39123 5.98436 6.14327 7.63337 7.84065 7.86015 7.87325 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OpenVINO This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Person Detection FP16 - Device: CPU Xeon Platinum 8592+ Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 EPYC 9654 2P Xeon Max 9480 2P EPYC 9754 2P Xeon Platinum 8490H EPYC 9684X Xeon Platinum 8380 2P EPYC 8534PN EPYC 9654 EPYC 8534P EPYC 9754 Xeon Platinum 8380 Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 60 120 180 240 300 SE +/- 0.09, N = 3 SE +/- 0.18, N = 3 SE +/- 0.10, N = 3 SE +/- 0.09, N = 3 SE +/- 0.09, N = 3 SE +/- 0.05, N = 3 SE +/- 0.10, N = 3 SE +/- 0.44, N = 15 SE +/- 0.06, N = 3 SE +/- 1.90, N = 15 SE +/- 0.01, N = 3 SE +/- 1.22, N = 15 SE +/- 0.14, N = 3 SE +/- 0.41, N = 3 SE +/- 2.90, N = 15 SE +/- 0.19, N = 3 SE +/- 6.97, N = 15 SE +/- 0.92, N = 3 SE +/- 0.15, N = 3 SE +/- 0.34, N = 3 SE +/- 0.25, N = 3 SE +/- 0.07, N = 3 39.62 41.93 42.16 43.41 43.77 45.78 59.69 63.98 64.91 85.86 100.39 110.91 113.70 156.59 157.45 161.48 230.24 236.81 266.91 274.52 283.83 293.02 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2024.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon Platinum 8592+ Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P Xeon Platinum 8592+ 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.08 / Avg: 138.77 / Max: 151.09 Min: 19.75 / Avg: 155.49 / Max: 170.23 Min: 34.28 / Avg: 219.81 / Max: 297.36 Min: 43.7 / Avg: 221.35 / Max: 243.97 Min: 51.68 / Avg: 253.57 / Max: 279.32 Min: 43.59 / Avg: 263.16 / Max: 290.9 Min: 20.51 / Avg: 267.71 / Max: 344.82 Min: 28.22 / Avg: 304.28 / Max: 347.63 Min: 59.06 / Avg: 327.33 / Max: 376.83 Min: 75.73 / Avg: 330.51 / Max: 371.12 Min: 74.3 / Avg: 331.92 / Max: 379.36 Min: 34.13 / Avg: 363.03 / Max: 408.76 Min: 104.28 / Avg: 438.97 / Max: 506.66 Min: 101.92 / Avg: 505.26 / Max: 553.92 Min: 106.87 / Avg: 515.24 / Max: 592.9 Min: 83.28 / Avg: 525.5 / Max: 710.63 Min: 40.79 / Avg: 578.55 / Max: 688.18 Min: 55.49 / Avg: 628.28 / Max: 705.07 Min: 166.26 / Avg: 657.43 / Max: 746.56 Min: 198.61 / Avg: 661.77 / Max: 760.61 Min: 197.25 / Avg: 664.45 / Max: 750.81 Min: 69.94 / Avg: 750.87 / Max: 819.03
OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Person Detection FP16 - Device: CPU EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X Xeon Platinum 8592+ EPYC 9654 Xeon Max 9480 Xeon Platinum 8490H Xeon Max 9468 EPYC 9754 Xeon 6780E 2P Xeon 6766E 2P EPYC 8534PN EPYC 8534P Xeon Platinum 8380 2P Xeon 6780E Xeon 6766E Xeon Platinum 8380 200 400 600 800 1000 SE +/- 2.17, N = 3 SE +/- 20.94, N = 15 SE +/- 5.47, N = 15 SE +/- 1.42, N = 3 SE +/- 0.69, N = 3 SE +/- 0.55, N = 3 SE +/- 1.28, N = 3 SE +/- 5.20, N = 15 SE +/- 0.95, N = 3 SE +/- 7.13, N = 15 SE +/- 0.52, N = 3 SE +/- 0.04, N = 3 SE +/- 1.23, N = 3 SE +/- 13.40, N = 15 SE +/- 0.34, N = 3 SE +/- 0.07, N = 3 SE +/- 0.56, N = 3 SE +/- 0.25, N = 3 SE +/- 0.23, N = 3 SE +/- 0.05, N = 3 SE +/- 0.09, N = 3 SE +/- 0.34, N = 3 1095.79 750.96 750.11 736.23 654.84 569.60 568.81 433.06 403.20 306.27 301.30 298.56 285.95 283.18 261.64 245.27 204.05 197.91 175.72 134.77 126.64 84.30 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: GPT-2 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9468 Xeon Max 9480 EPYC 9684X EPYC 9684X 2P Xeon 6780E Xeon Max 9468 2P Xeon 6766E Xeon 6766E 2P Xeon Max 9480 2P EPYC 9654 EPYC 9654 2P Xeon 6780E 2P EPYC 9754 EPYC 8534P EPYC 8534PN EPYC 9754 2P 3 6 9 12 15 SE +/- 0.01631, N = 3 SE +/- 0.15083, N = 15 SE +/- 0.03075, N = 3 SE +/- 0.04594, N = 3 SE +/- 0.08423, N = 15 SE +/- 0.03218, N = 3 SE +/- 0.06596, N = 15 SE +/- 0.06620, N = 3 SE +/- 0.02661, N = 3 SE +/- 0.01742, N = 3 SE +/- 0.02008, N = 3 SE +/- 0.02902, N = 3 SE +/- 0.02403, N = 3 SE +/- 0.02533, N = 3 SE +/- 0.06105, N = 3 SE +/- 0.08751, N = 5 SE +/- 0.01221, N = 3 SE +/- 0.80628, N = 15 SE +/- 0.01807, N = 3 SE +/- 0.03478, N = 3 SE +/- 0.07964, N = 3 SE +/- 0.20120, N = 15 3.69649 4.29531 4.49976 4.80418 5.16480 5.55101 5.67700 5.91961 6.16668 6.17190 6.74630 7.04111 7.16440 7.32610 7.53259 7.88584 9.74614 10.58571 11.13920 11.20470 11.28640 11.57740 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P 60 120 180 240 300 Min: 15.72 / Avg: 69.12 / Max: 74.02 Min: 16.03 / Avg: 72.08 / Max: 76.76 Min: 43.72 / Avg: 86.91 / Max: 91.46 Min: 43.54 / Avg: 90.3 / Max: 95.5 Min: 20.73 / Avg: 94.62 / Max: 100.76 Min: 27.83 / Avg: 127.32 / Max: 135.72 Min: 103.86 / Avg: 148.82 / Max: 155.68 Min: 34.75 / Avg: 149.45 / Max: 159.93 Min: 105.25 / Avg: 166.09 / Max: 183.4 Min: 52.71 / Avg: 203.23 / Max: 218.58 Min: 39.34 / Avg: 208.62 / Max: 222.38 Min: 55.76 / Avg: 257.41 / Max: 272.32 Min: 65.9 / Avg: 291.34 / Max: 308.52 Min: 97.22 / Avg: 292.43 / Max: 327.64
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: GPT-2 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9468 Xeon Max 9480 EPYC 9684X EPYC 9684X 2P Xeon 6780E Xeon Max 9468 2P Xeon 6766E Xeon 6766E 2P Xeon Max 9480 2P EPYC 9654 Xeon 6780E 2P EPYC 9654 2P EPYC 9754 EPYC 8534P EPYC 8534PN EPYC 9754 2P 60 120 180 240 300 SE +/- 1.19, N = 3 SE +/- 8.14, N = 15 SE +/- 1.50, N = 3 SE +/- 1.97, N = 3 SE +/- 3.16, N = 15 SE +/- 1.05, N = 3 SE +/- 2.01, N = 15 SE +/- 1.88, N = 3 SE +/- 0.70, N = 3 SE +/- 0.46, N = 3 SE +/- 0.44, N = 3 SE +/- 0.58, N = 3 SE +/- 0.47, N = 3 SE +/- 0.47, N = 3 SE +/- 1.08, N = 3 SE +/- 1.37, N = 5 SE +/- 8.91, N = 15 SE +/- 0.13, N = 3 SE +/- 0.15, N = 3 SE +/- 0.28, N = 3 SE +/- 0.63, N = 3 SE +/- 1.43, N = 15 270.33 236.65 222.03 207.93 194.15 180.03 176.33 168.86 162.09 161.95 148.11 141.95 139.45 136.39 132.71 126.82 103.89 102.57 89.74 89.21 88.57 86.69 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: GPT-2 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9468 Xeon Max 9480 EPYC 9684X EPYC 9684X 2P Xeon 6780E Xeon Max 9468 2P Xeon 6766E Xeon 6766E 2P Xeon Max 9480 2P EPYC 9654 EPYC 9654 2P Xeon 6780E 2P EPYC 9754 EPYC 8534P EPYC 8534PN EPYC 9754 2P 3 6 9 12 15 SE +/- 0.01631, N = 3 SE +/- 0.15083, N = 15 SE +/- 0.03075, N = 3 SE +/- 0.04594, N = 3 SE +/- 0.08423, N = 15 SE +/- 0.03218, N = 3 SE +/- 0.06596, N = 15 SE +/- 0.06620, N = 3 SE +/- 0.02661, N = 3 SE +/- 0.01742, N = 3 SE +/- 0.02008, N = 3 SE +/- 0.02902, N = 3 SE +/- 0.02403, N = 3 SE +/- 0.02533, N = 3 SE +/- 0.06105, N = 3 SE +/- 0.08751, N = 5 SE +/- 0.01221, N = 3 SE +/- 0.80628, N = 15 SE +/- 0.01807, N = 3 SE +/- 0.03478, N = 3 SE +/- 0.07964, N = 3 SE +/- 0.20120, N = 15 3.69649 4.29531 4.49976 4.80418 5.16480 5.55101 5.67700 5.91961 6.16668 6.17190 6.74630 7.04111 7.16440 7.32610 7.53259 7.88584 9.74614 10.58571 11.13920 11.20470 11.28640 11.57740 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 256 - Model: GoogLeNet Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P EPYC 9754 Xeon Max 9480 2P Xeon Platinum 8490H EPYC 9654 Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9754 2P EPYC 9684X 2P Xeon Max 9480 Xeon Max 9468 Xeon 6780E 2P EPYC 8534P Xeon 6766E 2P Xeon Platinum 8380 EPYC 8534PN Xeon 6780E Xeon 6766E 110 220 330 440 550 SE +/- 4.87, N = 3 SE +/- 2.68, N = 3 SE +/- 10.93, N = 12 SE +/- 4.80, N = 3 SE +/- 0.24, N = 3 SE +/- 3.87, N = 15 SE +/- 2.29, N = 3 SE +/- 2.30, N = 3 SE +/- 2.76, N = 15 SE +/- 0.69, N = 3 SE +/- 1.02, N = 3 SE +/- 1.11, N = 3 SE +/- 0.51, N = 3 SE +/- 1.91, N = 3 SE +/- 1.67, N = 3 SE +/- 2.49, N = 3 SE +/- 0.04, N = 3 SE +/- 3.73, N = 4 SE +/- 0.16, N = 3 SE +/- 0.07, N = 3 SE +/- 0.07, N = 3 SE +/- 0.31, N = 3 528.93 501.39 462.79 449.91 448.46 447.61 447.50 423.59 406.58 402.78 401.84 397.44 391.40 386.00 375.73 335.38 308.46 301.46 290.27 287.42 281.62 260.83 images/sec Per Watt
OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 256 - Model: GoogLeNet EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9684X Xeon 6766E Xeon Max 9480 Xeon 6780E Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8592+ 2P EPYC 9754 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon 6766E 2P Xeon 6780E 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 0.5108 1.0216 1.5324 2.0432 2.554 2.270 2.115 1.896 1.620 1.503 1.385 1.305 1.298 1.170 1.147 1.141 1.138 0.936 0.932 0.833 0.832 0.775 0.743 0.726 0.690 0.684 0.666 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.16.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E Xeon Platinum 8380 EPYC 9654 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 17.59 / Avg: 126.6 / Max: 137.81 Min: 18.95 / Avg: 145.84 / Max: 158.71 Min: 43.66 / Avg: 200.89 / Max: 222.56 Min: 21.26 / Avg: 244.09 / Max: 296.43 Min: 43.42 / Avg: 245.54 / Max: 272.59 Min: 53.8 / Avg: 255 / Max: 274.03 Min: 27.68 / Avg: 276.32 / Max: 312.83 Min: 33.46 / Avg: 281.92 / Max: 317.12 Min: 30.3 / Avg: 311.48 / Max: 350.99 Min: 58.65 / Avg: 323.79 / Max: 362.41 Min: 73.87 / Avg: 329.22 / Max: 366.78 Min: 79.85 / Avg: 329.98 / Max: 363.65 Min: 105.28 / Avg: 405.89 / Max: 469.79 Min: 40.92 / Avg: 426.33 / Max: 487.57 Min: 107.92 / Avg: 461.73 / Max: 537.33 Min: 103.67 / Avg: 483.98 / Max: 536.18 Min: 57.21 / Avg: 518.78 / Max: 576.81 Min: 82.2 / Avg: 535.58 / Max: 613.03 Min: 65.43 / Avg: 587.49 / Max: 655.55 Min: 161.35 / Avg: 635.04 / Max: 705.2 Min: 201.04 / Avg: 651.98 / Max: 714.73 Min: 201.04 / Avg: 653.82 / Max: 709.07
OSPRay Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Items Per Second, More Is Better OSPRay 3.1 Benchmark: particle_volume/ao/real_time EPYC 9754 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon 6780E 2P EPYC 9754 Xeon Max 9480 2P Xeon 6766E 2P EPYC 9684X Xeon Platinum 8592+ Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8490H Xeon 6780E Xeon 6766E Xeon Max 9480 EPYC 8534P Xeon Platinum 8380 2P EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 13 26 39 52 65 SE +/- 0.01800, N = 3 SE +/- 0.02899, N = 3 SE +/- 0.07203, N = 3 SE +/- 0.01406, N = 3 SE +/- 0.02269, N = 3 SE +/- 0.03595, N = 3 SE +/- 0.03039, N = 3 SE +/- 0.00297, N = 3 SE +/- 0.03817, N = 3 SE +/- 0.01754, N = 3 SE +/- 0.01144, N = 3 SE +/- 0.01945, N = 3 SE +/- 0.01424, N = 3 SE +/- 0.00975, N = 3 SE +/- 0.01225, N = 3 SE +/- 0.02488, N = 3 SE +/- 0.01050, N = 3 SE +/- 0.02190, N = 3 SE +/- 0.03319, N = 3 SE +/- 0.04390, N = 3 SE +/- 0.03513, N = 3 SE +/- 0.05948, N = 3 56.69360 55.33820 54.11950 47.61430 38.22350 36.50430 35.24610 33.11250 32.96630 32.51880 30.84040 29.50180 28.09560 21.93240 21.62220 19.51590 18.55330 17.96680 17.82690 16.45040 16.05020 9.77764 Items Per Second Per Watt
OpenBenchmarking.org Items Per Second Per Watt, More Is Better OSPRay 3.1 Benchmark: particle_volume/ao/real_time EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6766E EPYC 9654 EPYC 9754 2P Xeon 6780E Xeon Platinum 8592+ Xeon 6766E 2P EPYC 9684X Xeon Platinum 8592+ 2P Xeon 6780E 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 0.029 0.058 0.087 0.116 0.145 0.129 0.121 0.114 0.107 0.096 0.095 0.092 0.091 0.087 0.085 0.080 0.079 0.079 0.072 0.065 0.060 0.056 0.056 0.049 0.043 0.037 0.034 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OSPRay 3.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 Xeon Max 9468 EPYC 9654 EPYC 9754 Xeon Max 9480 Xeon Platinum 8490H Xeon Platinum 8592+ Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.71 / Avg: 127.12 / Max: 134.69 Min: 17.75 / Avg: 148.12 / Max: 155.49 Min: 43.7 / Avg: 183.09 / Max: 194.2 Min: 43.56 / Avg: 235.42 / Max: 250.81 Min: 53.36 / Avg: 261.49 / Max: 284.07 Min: 74.5 / Avg: 268.43 / Max: 329.84 Min: 28.62 / Avg: 291.85 / Max: 303.75 Min: 23.11 / Avg: 308.84 / Max: 321.58 Min: 78.47 / Avg: 334.1 / Max: 367.59 Min: 59.83 / Avg: 336.31 / Max: 366.79 Min: 35 / Avg: 338.48 / Max: 364.64 Min: 106.01 / Avg: 377.24 / Max: 400.35 Min: 36.05 / Avg: 383.1 / Max: 399.03 Min: 108.67 / Avg: 463.55 / Max: 488.13 Min: 103.11 / Avg: 521.41 / Max: 554.74 Min: 42.49 / Avg: 596.49 / Max: 627.65 Min: 60.35 / Avg: 600.41 / Max: 624.73 Min: 77.58 / Avg: 677.36 / Max: 715.65 Min: 195.3 / Avg: 681.19 / Max: 725.09 Min: 176.11 / Avg: 681.46 / Max: 714.09 Min: 200.34 / Avg: 682.67 / Max: 705.07 Min: 61.93 / Avg: 764.28 / Max: 797.27
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: yolov4 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9468 Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E EPYC 9654 EPYC 9684X EPYC 8534P EPYC 8534PN EPYC 9754 Xeon 6780E 2P Xeon 6766E 2P EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P 30 60 90 120 150 SE +/- 0.10, N = 3 SE +/- 0.38, N = 3 SE +/- 0.12, N = 3 SE +/- 0.55, N = 15 SE +/- 0.35, N = 3 SE +/- 0.55, N = 15 SE +/- 0.25, N = 3 SE +/- 0.28, N = 3 SE +/- 0.47, N = 14 SE +/- 1.43, N = 15 SE +/- 0.51, N = 3 SE +/- 0.37, N = 3 SE +/- 0.88, N = 8 SE +/- 1.34, N = 3 SE +/- 1.36, N = 4 SE +/- 1.32, N = 5 SE +/- 0.77, N = 3 SE +/- 0.25, N = 3 SE +/- 1.08, N = 3 SE +/- 1.60, N = 3 SE +/- 2.08, N = 3 SE +/- 1.03, N = 3 43.54 49.39 50.35 54.96 58.94 59.18 60.93 64.29 69.20 69.45 72.31 78.73 101.47 102.73 112.73 114.31 120.56 122.75 133.36 137.99 148.21 155.96 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P 110 220 330 440 550 Min: 16.87 / Avg: 112.11 / Max: 123.24 Min: 18.9 / Avg: 123.51 / Max: 135.75 Min: 43.64 / Avg: 163.17 / Max: 174.94 Min: 20.25 / Avg: 189.72 / Max: 207.12 Min: 43.43 / Avg: 192.41 / Max: 207.48 Min: 25.1 / Avg: 234.83 / Max: 253.99 Min: 48.27 / Avg: 242.83 / Max: 266.43 Min: 34.28 / Avg: 276.26 / Max: 300.81 Min: 103.87 / Avg: 332.75 / Max: 361.02 Min: 105.31 / Avg: 380 / Max: 411.37 Min: 40.24 / Avg: 404.21 / Max: 436.31 Min: 98.37 / Avg: 446.63 / Max: 480.06 Min: 60.04 / Avg: 479.74 / Max: 516.86 Min: 67.43 / Avg: 546.46 / Max: 595.03
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: yolov4 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9468 Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon 6780E Xeon 6766E EPYC 9654 EPYC 9684X EPYC 8534P EPYC 8534PN EPYC 9754 Xeon 6780E 2P Xeon 6766E 2P EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P 6 12 18 24 30 SE +/- 0.05084, N = 3 SE +/- 0.15605, N = 3 SE +/- 0.04802, N = 3 SE +/- 0.16821, N = 15 SE +/- 0.09991, N = 3 SE +/- 0.15263, N = 15 SE +/- 0.06818, N = 3 SE +/- 0.06825, N = 3 SE +/- 0.29451, N = 15 SE +/- 0.09595, N = 14 SE +/- 0.09768, N = 3 SE +/- 0.05969, N = 3 SE +/- 0.08554, N = 8 SE +/- 0.12519, N = 3 SE +/- 0.10588, N = 4 SE +/- 0.09781, N = 5 SE +/- 0.05314, N = 3 SE +/- 0.01654, N = 3 SE +/- 0.06011, N = 3 SE +/- 0.08454, N = 3 SE +/- 0.09575, N = 3 SE +/- 0.04207, N = 3 22.96890 20.24950 19.85930 18.21790 16.96720 16.91660 16.41230 15.55550 14.48310 14.45830 13.82950 12.70150 9.86018 9.73685 8.87446 8.75233 8.29533 8.14615 7.49940 7.24869 6.74977 6.41229 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: yolov4 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9468 Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E EPYC 9654 EPYC 9684X EPYC 8534P EPYC 8534PN EPYC 9754 Xeon 6780E 2P Xeon 6766E 2P EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P 30 60 90 120 150 SE +/- 0.10, N = 3 SE +/- 0.38, N = 3 SE +/- 0.12, N = 3 SE +/- 0.55, N = 15 SE +/- 0.35, N = 3 SE +/- 0.55, N = 15 SE +/- 0.25, N = 3 SE +/- 0.28, N = 3 SE +/- 0.47, N = 14 SE +/- 1.43, N = 15 SE +/- 0.51, N = 3 SE +/- 0.37, N = 3 SE +/- 0.88, N = 8 SE +/- 1.34, N = 3 SE +/- 1.36, N = 4 SE +/- 1.32, N = 5 SE +/- 0.77, N = 3 SE +/- 0.25, N = 3 SE +/- 1.08, N = 3 SE +/- 1.60, N = 3 SE +/- 2.08, N = 3 SE +/- 1.03, N = 3 43.54 49.39 50.35 54.96 58.94 59.18 60.93 64.29 69.20 69.45 72.31 78.73 101.47 102.73 112.73 114.31 120.56 122.75 133.36 137.99 148.21 155.96 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: super-resolution-10 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8380 Xeon Platinum 8380 2P EPYC 9684X EPYC 9654 EPYC 9684X 2P EPYC 8534P EPYC 9754 EPYC 8534PN EPYC 9754 2P EPYC 9654 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E 2P Xeon 6766E 4 8 12 16 20 SE +/- 0.00044, N = 3 SE +/- 0.05342, N = 3 SE +/- 0.00313, N = 3 SE +/- 0.00187, N = 3 SE +/- 0.00532, N = 3 SE +/- 0.10855, N = 15 SE +/- 0.05347, N = 4 SE +/- 0.02558, N = 3 SE +/- 0.04501, N = 3 SE +/- 0.28946, N = 12 SE +/- 0.03798, N = 3 SE +/- 0.04482, N = 3 SE +/- 0.18371, N = 12 SE +/- 0.02616, N = 3 SE +/- 0.02331, N = 3 SE +/- 0.02847, N = 3 SE +/- 0.26893, N = 15 SE +/- 0.06617, N = 3 SE +/- 0.07241, N = 3 SE +/- 0.14238, N = 15 SE +/- 0.07602, N = 3 SE +/- 0.08782, N = 3 3.85774 3.98013 4.39649 4.58912 4.67269 4.72530 4.85818 5.00222 5.05427 5.29844 8.13326 8.15092 8.39969 9.38681 9.42177 9.71131 10.23106 10.43440 15.10570 15.12740 16.80620 16.86310 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 EPYC 9684X EPYC 9754 2P Xeon Platinum 8380 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P 60 120 180 240 300 Min: 33.7 / Avg: 65.57 / Max: 67.97 Min: 43.48 / Avg: 72.93 / Max: 75.76 Min: 16.93 / Avg: 76.89 / Max: 81.74 Min: 18.73 / Avg: 83.53 / Max: 88.8 Min: 21.68 / Avg: 106.78 / Max: 113.35 Min: 103.86 / Avg: 127.8 / Max: 131.15 Min: 104.68 / Avg: 138.22 / Max: 146.78 Min: 27.15 / Avg: 144.22 / Max: 153.32 Min: 28.38 / Avg: 160.26 / Max: 175.87 Min: 40.2 / Avg: 219.21 / Max: 234.95 Min: 51.72 / Avg: 236.59 / Max: 257.65 Min: 57.88 / Avg: 270.43 / Max: 286.65 Min: 62.01 / Avg: 306.76 / Max: 325.52 Min: 98.64 / Avg: 328.39 / Max: 359.38
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: super-resolution-10 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8380 Xeon Platinum 8380 2P EPYC 9684X EPYC 9654 EPYC 9684X 2P EPYC 8534P EPYC 9754 EPYC 8534PN EPYC 9754 2P EPYC 9654 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E 2P Xeon 6766E 60 120 180 240 300 SE +/- 0.03, N = 3 SE +/- 3.39, N = 3 SE +/- 0.16, N = 3 SE +/- 0.09, N = 3 SE +/- 0.25, N = 3 SE +/- 4.15, N = 15 SE +/- 2.30, N = 4 SE +/- 1.02, N = 3 SE +/- 1.75, N = 3 SE +/- 6.83, N = 12 SE +/- 0.58, N = 3 SE +/- 0.67, N = 3 SE +/- 2.15, N = 12 SE +/- 0.30, N = 3 SE +/- 0.26, N = 3 SE +/- 0.30, N = 3 SE +/- 2.46, N = 15 SE +/- 0.61, N = 3 SE +/- 0.32, N = 3 SE +/- 0.60, N = 15 SE +/- 0.27, N = 3 SE +/- 0.31, N = 3 259.19 251.31 227.42 217.88 213.99 212.92 205.89 199.90 197.85 192.81 122.95 122.68 119.56 106.53 106.13 102.97 98.64 95.84 66.20 66.18 59.50 59.30 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: super-resolution-10 - Device: CPU - Executor: Standard Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8380 Xeon Platinum 8380 2P EPYC 9684X EPYC 9654 EPYC 9684X 2P EPYC 8534P EPYC 9754 EPYC 8534PN EPYC 9754 2P EPYC 9654 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E 2P Xeon 6766E 4 8 12 16 20 SE +/- 0.00044, N = 3 SE +/- 0.05342, N = 3 SE +/- 0.00313, N = 3 SE +/- 0.00187, N = 3 SE +/- 0.00532, N = 3 SE +/- 0.10855, N = 15 SE +/- 0.05347, N = 4 SE +/- 0.02558, N = 3 SE +/- 0.04501, N = 3 SE +/- 0.28946, N = 12 SE +/- 0.03798, N = 3 SE +/- 0.04482, N = 3 SE +/- 0.18371, N = 12 SE +/- 0.02616, N = 3 SE +/- 0.02331, N = 3 SE +/- 0.02847, N = 3 SE +/- 0.26893, N = 15 SE +/- 0.06617, N = 3 SE +/- 0.07241, N = 3 SE +/- 0.14238, N = 15 SE +/- 0.07602, N = 3 SE +/- 0.08782, N = 3 3.85774 3.98013 4.39649 4.58912 4.67269 4.72530 4.85818 5.00222 5.05427 5.29844 8.13326 8.15092 8.39969 9.38681 9.42177 9.71131 10.23106 10.43440 15.10570 15.12740 16.80620 16.86310 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OSPRay Studio Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 1.0 Camera: 1 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPU EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9754 EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon 6780E 2P Xeon Platinum 8490H Xeon 6766E 2P Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon 6780E Xeon 6766E Xeon Platinum 8380 20K 40K 60K 80K 100K SE +/- 47.34, N = 3 SE +/- 14.62, N = 3 SE +/- 109.43, N = 3 SE +/- 16.05, N = 3 SE +/- 64.24, N = 3 SE +/- 13.86, N = 3 SE +/- 312.83, N = 12 SE +/- 193.23, N = 3 SE +/- 37.55, N = 3 SE +/- 42.05, N = 3 SE +/- 90.37, N = 3 SE +/- 152.29, N = 3 SE +/- 605.82, N = 15 SE +/- 102.40, N = 3 SE +/- 739.02, N = 15 SE +/- 64.91, N = 3 SE +/- 134.33, N = 3 SE +/- 172.59, N = 3 SE +/- 346.19, N = 3 SE +/- 424.84, N = 3 SE +/- 541.85, N = 3 SE +/- 175.68, N = 3 16960 17855 20750 24249 33411 36179 37383 38645 38926 43290 50798 55247 56082 58026 62449 62540 68973 72051 78093 81003 88579 103461 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 1.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9754 2P Xeon Platinum 8490H 2P EPYC 9654 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.4 / Avg: 117.68 / Max: 142.04 Min: 17.98 / Avg: 135.52 / Max: 167.31 Min: 43.65 / Avg: 164.56 / Max: 223.84 Min: 43.57 / Avg: 204.48 / Max: 283.69 Min: 21.62 / Avg: 219.88 / Max: 335.09 Min: 54.98 / Avg: 238.47 / Max: 291.29 Min: 30.06 / Avg: 244.64 / Max: 330.87 Min: 35.11 / Avg: 281.84 / Max: 376.97 Min: 30.5 / Avg: 290.44 / Max: 410.63 Min: 60.65 / Avg: 292.18 / Max: 377.6 Min: 78.44 / Avg: 306.37 / Max: 375.68 Min: 69.88 / Avg: 308.64 / Max: 375.21 Min: 104.78 / Avg: 311.45 / Max: 486.78 Min: 107.48 / Avg: 352.84 / Max: 587.47 Min: 102.98 / Avg: 432.31 / Max: 581.57 Min: 42.88 / Avg: 522.41 / Max: 691.8 Min: 168.48 / Avg: 535.88 / Max: 754.67 Min: 60.75 / Avg: 559.19 / Max: 686.72 Min: 197.25 / Avg: 571.94 / Max: 754.03 Min: 213.26 / Avg: 577.01 / Max: 753.84 Min: 79.71 / Avg: 606.49 / Max: 752.52 Min: 69.54 / Avg: 653.11 / Max: 822.49
Result
OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 1.0 Camera: 3 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPU EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9754 EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon 6780E 2P Xeon Platinum 8490H Xeon 6766E 2P Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon 6780E Xeon 6766E Xeon Platinum 8380 30K 60K 90K 120K 150K SE +/- 28.11, N = 3 SE +/- 27.74, N = 3 SE +/- 756.08, N = 15 SE +/- 47.16, N = 3 SE +/- 125.33, N = 3 SE +/- 141.04, N = 3 SE +/- 50.97, N = 3 SE +/- 89.57, N = 3 SE +/- 155.80, N = 3 SE +/- 231.95, N = 3 SE +/- 162.52, N = 3 SE +/- 32.38, N = 3 SE +/- 176.25, N = 3 SE +/- 58.68, N = 3 SE +/- 900.17, N = 15 SE +/- 65.97, N = 3 SE +/- 41.33, N = 3 SE +/- 29.36, N = 3 SE +/- 13.58, N = 3 SE +/- 189.03, N = 3 SE +/- 365.52, N = 3 SE +/- 623.29, N = 3 20071 20893 27943 33391 38813 41971 42945 44349 45211 50019 59522 64188 66487 68006 71583 73366 80742 84261 91643 92175 101444 121734 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 1.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9684X Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.93 / Avg: 120.39 / Max: 142.49 Min: 18.47 / Avg: 138.95 / Max: 167.41 Min: 43.64 / Avg: 170.07 / Max: 223.38 Min: 43.57 / Avg: 212.15 / Max: 289.98 Min: 21.87 / Avg: 229.47 / Max: 334.93 Min: 52.23 / Avg: 243.3 / Max: 290.85 Min: 27.71 / Avg: 254.04 / Max: 331.67 Min: 34.02 / Avg: 290.6 / Max: 377.55 Min: 59.95 / Avg: 299.76 / Max: 377.1 Min: 34.85 / Avg: 304.67 / Max: 411.05 Min: 74.4 / Avg: 312.09 / Max: 377.67 Min: 69.72 / Avg: 313.73 / Max: 377.15 Min: 56.05 / Avg: 325.96 / Max: 489.46 Min: 107.91 / Avg: 369.36 / Max: 547.58 Min: 105.22 / Avg: 444.93 / Max: 582.81 Min: 81.52 / Avg: 518.28 / Max: 753.01 Min: 60.79 / Avg: 534.58 / Max: 678.46 Min: 43.47 / Avg: 541.66 / Max: 691.2 Min: 167.48 / Avg: 554.33 / Max: 756.36 Min: 194.1 / Avg: 583.5 / Max: 754.71 Min: 204.64 / Avg: 589.92 / Max: 755.25 Min: 68.36 / Avg: 672.33 / Max: 821.45
Zstd Compression This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.4 Compression Level: 19 - Decompression Speed EPYC 9684X 2P EPYC 9654 EPYC 9684X EPYC 9654 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 8534PN EPYC 9754 EPYC 9754 2P Xeon Max 9480 Xeon Max 9468 EPYC 8534P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8380 Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 300 600 900 1200 1500 SE +/- 1.73, N = 3 SE +/- 0.46, N = 3 SE +/- 7.31, N = 3 SE +/- 2.20, N = 3 SE +/- 3.01, N = 4 SE +/- 1.88, N = 3 SE +/- 1.58, N = 3 SE +/- 0.83, N = 3 SE +/- 1.83, N = 3 SE +/- 1.08, N = 15 SE +/- 2.12, N = 3 SE +/- 21.23, N = 3 SE +/- 0.66, N = 15 SE +/- 0.65, N = 15 SE +/- 1.75, N = 3 SE +/- 4.45, N = 3 SE +/- 4.96, N = 3 SE +/- 1.74, N = 4 SE +/- 5.13, N = 3 SE +/- 2.25, N = 3 SE +/- 2.22, N = 3 SE +/- 5.06, N = 3 1392.5 1388.9 1385.3 1382.1 1221.8 1165.8 1144.8 1139.9 1139.0 1127.5 1126.4 1118.8 1040.2 1039.2 1032.4 1032.0 963.6 900.5 816.7 774.4 743.5 721.4 1. (CC) gcc options: -O3 -pthread -lz -llzma MB/s Per Watt
OpenBenchmarking.org MB/s Per Watt, More Is Better Zstd Compression 1.5.4 Compression Level: 19 - Decompression Speed EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6780E EPYC 9654 Xeon 6766E EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6780E 2P EPYC 9754 2P Xeon 6766E 2P EPYC 9654 2P Xeon Max 9468 Xeon Max 9480 EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 5 10 15 20 25 20.546 19.698 14.065 13.679 13.308 12.746 11.711 9.439 8.705 6.465 6.162 5.875 5.845 5.818 5.616 5.548 5.292 4.454 4.008 2.828 2.520 2.506 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Zstd Compression 1.5.4 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Platinum 8380 2P Xeon Max 9480 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 80 160 240 320 400 Min: 16.36 / Avg: 55.72 / Max: 65.47 Min: 16.44 / Avg: 56.8 / Max: 67.51 Min: 43.62 / Avg: 58.33 / Max: 74.24 Min: 43.45 / Avg: 59.7 / Max: 75.66 Min: 21.5 / Avg: 81.05 / Max: 92.77 Min: 27.93 / Avg: 104.36 / Max: 120.51 Min: 50.67 / Avg: 109.38 / Max: 135.43 Min: 33.68 / Avg: 118.29 / Max: 136.36 Min: 103.62 / Avg: 123.42 / Max: 157.28 Min: 104.87 / Avg: 125.62 / Max: 159.28 Min: 33.62 / Avg: 140.35 / Max: 172.45 Min: 61.06 / Avg: 159.64 / Max: 183.39 Min: 38.29 / Avg: 193.87 / Max: 209.99 Min: 73.69 / Avg: 200.55 / Max: 225.84 Min: 98.06 / Avg: 202.2 / Max: 229.76 Min: 75.37 / Avg: 203.22 / Max: 228.29 Min: 59.25 / Avg: 237.56 / Max: 259.24 Min: 60.19 / Avg: 263.13 / Max: 287.32 Min: 76.44 / Avg: 290.84 / Max: 334.48 Min: 170.25 / Avg: 340.73 / Max: 371.6 Min: 197.43 / Avg: 412.33 / Max: 448.69 Min: 203.74 / Avg: 415.14 / Max: 451.39
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.4 Compression Level: 19 - Compression Speed Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9654 EPYC 9684X EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P EPYC 8534P EPYC 8534PN EPYC 9754 Xeon Max 9468 2P EPYC 9754 2P Xeon Platinum 8380 Xeon Max 9480 2P Xeon 6780E 2P Xeon 6780E Xeon 6766E Xeon 6766E 2P 4 8 12 16 20 SE +/- 0.21, N = 4 SE +/- 0.03, N = 3 SE +/- 0.00, N = 3 SE +/- 0.03, N = 3 SE +/- 0.09, N = 3 SE +/- 0.22, N = 3 SE +/- 0.12, N = 3 SE +/- 0.06, N = 3 SE +/- 0.11, N = 15 SE +/- 0.15, N = 3 SE +/- 0.17, N = 4 SE +/- 0.00, N = 3 SE +/- 0.03, N = 3 SE +/- 0.06, N = 3 SE +/- 0.09, N = 15 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.13, N = 15 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.00, N = 3 SE +/- 0.07, N = 3 18.0 17.8 17.2 17.2 17.0 16.8 15.2 14.6 14.5 14.4 14.3 13.9 13.8 13.8 13.8 13.7 13.7 13.3 11.6 11.5 10.6 10.6 1. (CC) gcc options: -O3 -pthread -lz -llzma
Speedb Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Op/s, More Is Better Speedb 2.7 Test: Random Read Xeon 6780E 2P EPYC 9754 2P Xeon 6766E 2P EPYC 9684X 2P EPYC 9654 2P Xeon 6780E Xeon 6766E EPYC 9754 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9480 2P EPYC 9654 Xeon Max 9468 2P Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 8534P Xeon Platinum 8380 2P Xeon Max 9480 EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 300M 600M 900M 1200M 1500M SE +/- 2674444.99, N = 3 SE +/- 2399164.12, N = 3 SE +/- 1272758.30, N = 3 SE +/- 5954964.38, N = 3 SE +/- 955915.20, N = 3 SE +/- 1322733.50, N = 3 SE +/- 998087.60, N = 3 SE +/- 1606748.01, N = 3 SE +/- 5399265.88, N = 15 SE +/- 5393159.45, N = 3 SE +/- 205170.43, N = 3 SE +/- 2955608.21, N = 3 SE +/- 66268.38, N = 3 SE +/- 8491098.40, N = 15 SE +/- 242146.75, N = 3 SE +/- 168680.90, N = 3 SE +/- 109015.29, N = 3 SE +/- 706582.63, N = 3 SE +/- 2763560.64, N = 13 SE +/- 70537.88, N = 3 SE +/- 3698537.65, N = 15 SE +/- 232538.80, N = 3 1324163340 1169947385 1158496546 1059546533 857687179 673159217 596955494 589951070 586147871 551286958 537633034 497907490 428557054 419943570 315178381 276179617 274091778 259136466 244070674 242940110 211457048 130359145 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread Op/s Per Watt
OpenBenchmarking.org Op/s Per Watt, More Is Better Speedb 2.7 Test: Random Read Xeon 6766E Xeon 6766E 2P Xeon 6780E Xeon 6780E 2P EPYC 8534PN EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 9654 EPYC 9654 2P EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9480 Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 500K 1000K 1500K 2000K 2500K 2531402.05 2449964.81 2159153.28 2130388.24 2023204.53 1991742.68 1962564.72 1818446.43 1563545.14 1511464.50 1404496.49 1380809.63 962315.70 894980.08 834970.47 834394.35 747196.35 744809.12 680456.82 644065.87 509439.26 506592.64 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Speedb 2.7 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 EPYC 9654 EPYC 9754 Xeon Max 9468 Xeon 6780E Xeon Platinum 8592+ Xeon Max 9480 Xeon Platinum 8490H EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E 2P Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 18.13 / Avg: 120.08 / Max: 129.07 Min: 18.42 / Avg: 150.73 / Max: 161.26 Min: 43.69 / Avg: 235.82 / Max: 270.43 Min: 54.55 / Avg: 255.89 / Max: 290.77 Min: 29.2 / Avg: 274.09 / Max: 293.61 Min: 22.11 / Avg: 296.2 / Max: 318.84 Min: 71.79 / Avg: 310.76 / Max: 333.42 Min: 43.47 / Avg: 311.77 / Max: 356.67 Min: 27.16 / Avg: 327.52 / Max: 377.21 Min: 70.78 / Avg: 327.7 / Max: 377.13 Min: 54.25 / Avg: 330.77 / Max: 378.66 Min: 33.89 / Avg: 382.79 / Max: 411.07 Min: 105.77 / Avg: 472.86 / Max: 540.08 Min: 103.26 / Avg: 511.53 / Max: 582.64 Min: 60.4 / Avg: 567.45 / Max: 609.28 Min: 40.77 / Avg: 596.13 / Max: 648.34 Min: 107.84 / Avg: 621.56 / Max: 712.89 Min: 185.93 / Avg: 652.02 / Max: 753.16 Min: 68.16 / Avg: 654.93 / Max: 718.87 Min: 168.96 / Avg: 660.7 / Max: 755.23 Min: 196.64 / Avg: 666.37 / Max: 752.39 Min: 69.56 / Avg: 767.34 / Max: 822.12
VVenC VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better VVenC 1.11 Video Input: Bosphorus 4K - Video Preset: Fast EPYC 9684X EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8490H 2P EPYC 9754 Xeon Max 9468 EPYC 8534P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 8534PN Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon 6780E Xeon 6766E 2 4 6 8 10 SE +/- 0.057, N = 3 SE +/- 0.048, N = 3 SE +/- 0.023, N = 3 SE +/- 0.038, N = 3 SE +/- 0.022, N = 3 SE +/- 0.033, N = 3 SE +/- 0.047, N = 3 SE +/- 0.045, N = 3 SE +/- 0.010, N = 3 SE +/- 0.038, N = 3 SE +/- 0.043, N = 15 SE +/- 0.081, N = 3 SE +/- 0.041, N = 3 SE +/- 0.028, N = 3 SE +/- 0.010, N = 3 SE +/- 0.007, N = 3 SE +/- 0.008, N = 3 7.547 7.431 7.392 6.960 6.719 6.501 6.480 6.465 6.324 6.118 6.077 6.023 6.000 5.877 5.583 4.440 4.052 1. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better VVenC 1.11 Video Input: Bosphorus 4K - Video Preset: Fast EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.0151 0.0302 0.0453 0.0604 0.0755 0.067 0.062 0.052 0.050 0.048 0.042 0.037 0.032 0.028 0.028 0.023 0.023 0.018 0.017 0.015 0.012 0.012 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better VVenC 1.11 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 110 220 330 440 550 Min: 43.59 / Avg: 78.47 / Max: 122.17 Min: 43.46 / Avg: 89.13 / Max: 146.7 Min: 15.3 / Avg: 89.82 / Max: 125.64 Min: 16.24 / Avg: 99.45 / Max: 145.04 Min: 20.88 / Avg: 134.72 / Max: 195.13 Min: 27.21 / Avg: 177.95 / Max: 263.84 Min: 52.86 / Avg: 196.6 / Max: 265.58 Min: 31.5 / Avg: 204.88 / Max: 302.64 Min: 26.5 / Avg: 232.82 / Max: 325.56 Min: 59.59 / Avg: 242.83 / Max: 330.87 Min: 71.41 / Avg: 273.3 / Max: 339.11 Min: 84.83 / Avg: 277.48 / Max: 349.11 Min: 100.2 / Avg: 328.02 / Max: 463.25 Min: 77.27 / Avg: 413.94 / Max: 550.57 Min: 154.77 / Avg: 444.48 / Max: 576.57 Min: 218.19 / Avg: 515.79 / Max: 625.23 Min: 193.59 / Avg: 516.46 / Max: 637.28 Video Input: Bosphorus 4K - Video Preset: Fast
EPYC 9684X 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1
EPYC 9654 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1
EPYC 9754 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1
Xeon 6766E 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1
Xeon 6780E 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1
Zstd Compression This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.4 Compression Level: 19, Long Mode - Decompression Speed EPYC 9654 EPYC 9684X EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Max 9468 EPYC 9754 2P EPYC 9754 EPYC 8534P EPYC 8534PN Xeon Platinum 8490H Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 300 600 900 1200 1500 SE +/- 1.17, N = 3 SE +/- 1.99, N = 3 SE +/- 1.87, N = 3 SE +/- 2.93, N = 3 SE +/- 0.84, N = 3 SE +/- 1.60, N = 3 SE +/- 1.45, N = 3 SE +/- 0.56, N = 3 SE +/- 0.19, N = 3 SE +/- 1.95, N = 3 SE +/- 2.09, N = 3 SE +/- 0.92, N = 3 SE +/- 0.43, N = 3 SE +/- 1.78, N = 15 SE +/- 1.00, N = 3 SE +/- 4.15, N = 3 SE +/- 3.69, N = 3 SE +/- 0.30, N = 3 SE +/- 1.05, N = 3 SE +/- 1.54, N = 3 SE +/- 3.99, N = 3 SE +/- 4.61, N = 3 1328.5 1324.9 1324.6 1321.4 1242.6 1189.7 1133.3 1131.1 1104.2 1102.2 1101.4 1096.5 1069.1 1046.0 1041.8 1018.3 992.1 912.9 797.7 753.2 739.1 701.2 1. (CC) gcc options: -O3 -pthread -lz -llzma MB/s Per Watt
OpenBenchmarking.org MB/s Per Watt, More Is Better Zstd Compression 1.5.4 Compression Level: 19, Long Mode - Decompression Speed EPYC 8534PN EPYC 8534P Xeon 6780E Xeon 6766E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 5 10 15 20 25 20.797 20.514 15.034 14.311 14.042 13.268 11.661 9.833 9.241 6.931 6.333 6.001 5.816 5.787 5.674 5.643 5.089 4.675 4.173 2.956 2.539 2.512 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Zstd Compression 1.5.4 CPU Power Consumption Monitor Xeon 6766E EPYC 8534PN Xeon 6780E EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Platinum 8380 2P Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 80 160 240 320 400 Min: 43.63 / Avg: 51.65 / Max: 69.7 Min: 15.43 / Avg: 52.72 / Max: 59.1 Min: 43.37 / Avg: 53.06 / Max: 71.35 Min: 17.05 / Avg: 53.69 / Max: 60.18 Min: 18.92 / Avg: 78.5 / Max: 87.36 Min: 28.35 / Avg: 100.13 / Max: 109.97 Min: 49.33 / Avg: 103.56 / Max: 117.6 Min: 33.27 / Avg: 113.62 / Max: 124.55 Min: 103.71 / Avg: 116.84 / Max: 150.87 Min: 105.69 / Avg: 118.94 / Max: 149.49 Min: 33.99 / Avg: 134.46 / Max: 153.43 Min: 61.25 / Avg: 154.25 / Max: 170.09 Min: 39.89 / Avg: 189.86 / Max: 203.51 Min: 96.44 / Avg: 195.29 / Max: 212.05 Min: 73.9 / Avg: 195.37 / Max: 210.14 Min: 82.48 / Avg: 199.75 / Max: 214.36 Min: 57.79 / Avg: 234.74 / Max: 249.39 Min: 64.16 / Avg: 259.66 / Max: 276.36 Min: 75.89 / Avg: 285.1 / Max: 313.97 Min: 172.05 / Avg: 335.67 / Max: 356.47 Min: 201.8 / Avg: 412.05 / Max: 436.43 Min: 208.83 / Avg: 414.8 / Max: 438.11
OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.4 Compression Level: 19, Long Mode - Compression Speed Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9684X 2P EPYC 9654 EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon Max 9468 Xeon Max 9480 Xeon Max 9468 2P Xeon Platinum 8380 Xeon Max 9480 2P EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9754 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 3 6 9 12 15 SE +/- 0.05, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.06, N = 15 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 SE +/- 0.00, N = 3 SE +/- 0.04, N = 3 9.80 9.67 8.45 8.42 8.38 8.30 8.16 7.94 7.83 7.76 7.72 7.62 7.52 7.46 6.76 6.74 6.68 6.64 6.19 6.10 5.65 5.60 1. (CC) gcc options: -O3 -pthread -lz -llzma
OpenVINO This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 EPYC 9684X 2P EPYC 9654 2P Xeon Max 9480 EPYC 9754 2P Xeon Platinum 8490H Xeon Platinum 8380 2P EPYC 9684X Xeon Platinum 8592+ 2P EPYC 9654 Xeon Platinum 8592+ EPYC 8534P Xeon Platinum 8380 EPYC 8534PN EPYC 9754 Xeon 6780E 2P Xeon 6766E 2P Xeon 6780E Xeon 6766E 0.306 0.612 0.918 1.224 1.53 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 15 SE +/- 0.00, N = 3 SE +/- 0.00, N = 7 SE +/- 0.00, N = 15 SE +/- 0.01, N = 3 SE +/- 0.00, N = 9 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 0.28 0.29 0.30 0.33 0.35 0.35 0.37 0.41 0.43 0.45 0.45 0.46 0.48 0.49 0.53 0.56 0.60 0.73 1.09 1.20 1.25 1.36 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2024.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8592+ Xeon 6780E Xeon Platinum 8380 Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 EPYC 9654 Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9684X EPYC 9754 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9480 2P Xeon Max 9468 2P 120 240 360 480 600 Min: 16.42 / Avg: 129.12 / Max: 139.81 Min: 19.35 / Avg: 152.78 / Max: 164.84 Min: 43.65 / Avg: 203.6 / Max: 221.29 Min: 33.52 / Avg: 235.56 / Max: 293.33 Min: 43.63 / Avg: 241.33 / Max: 263.98 Min: 51.57 / Avg: 255.35 / Max: 290.92 Min: 104.83 / Avg: 274.08 / Max: 295.37 Min: 107.98 / Avg: 302.19 / Max: 331.56 Min: 21.53 / Avg: 310.47 / Max: 338.08 Min: 26.97 / Avg: 310.58 / Max: 341.7 Min: 59.31 / Avg: 328.19 / Max: 373.29 Min: 73.52 / Avg: 330.72 / Max: 377.05 Min: 83.36 / Avg: 331.78 / Max: 377.93 Min: 34.06 / Avg: 367.01 / Max: 400.16 Min: 40.6 / Avg: 436.99 / Max: 487.81 Min: 81.83 / Avg: 472.9 / Max: 559.21 Min: 104.06 / Avg: 507.52 / Max: 580.05 Min: 57.78 / Avg: 545.61 / Max: 614.54 Min: 169.81 / Avg: 585.54 / Max: 631.87 Min: 36.23 / Avg: 623.56 / Max: 689.59 Min: 200.92 / Avg: 627.67 / Max: 669.86 Min: 201.96 / Avg: 633.47 / Max: 674.77
OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon Max 9480 EPYC 9684X Xeon Max 9468 EPYC 9654 EPYC 9754 Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon 6780E EPYC 8534P Xeon 6766E Xeon Platinum 8592+ EPYC 8534PN Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8380 40K 80K 120K 160K 200K SE +/- 2494.14, N = 3 SE +/- 1640.61, N = 7 SE +/- 1422.88, N = 15 SE +/- 18.38, N = 3 SE +/- 1922.38, N = 15 SE +/- 1374.23, N = 9 SE +/- 696.52, N = 3 SE +/- 1761.33, N = 3 SE +/- 162.47, N = 3 SE +/- 383.22, N = 3 SE +/- 205.87, N = 3 SE +/- 89.35, N = 3 SE +/- 1158.60, N = 3 SE +/- 75.30, N = 3 SE +/- 100.17, N = 3 SE +/- 86.99, N = 3 SE +/- 68.06, N = 3 SE +/- 988.12, N = 3 SE +/- 55.52, N = 3 SE +/- 1067.91, N = 3 SE +/- 1048.23, N = 3 SE +/- 95.54, N = 3 189893.31 178488.03 176900.32 174862.56 172688.50 166151.42 151627.42 140144.01 138605.52 135075.07 133779.62 132973.43 128566.81 115106.30 91273.02 87511.11 84584.13 81279.79 78334.92 76895.29 75761.50 61956.93 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OSPRay Studio Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 1.0 Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPU EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9754 Xeon Max 9480 2P EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon 6780E 2P Xeon Platinum 8592+ Xeon 6766E 2P Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 8534P Xeon 6780E EPYC 8534PN Xeon 6766E Xeon Platinum 8380 700 1400 2100 2800 3500 SE +/- 0.88, N = 3 SE +/- 1.20, N = 3 SE +/- 0.58, N = 3 SE +/- 0.88, N = 3 SE +/- 0.58, N = 3 SE +/- 1.20, N = 3 SE +/- 0.58, N = 3 SE +/- 1.15, N = 3 SE +/- 1.76, N = 3 SE +/- 0.58, N = 3 SE +/- 24.01, N = 12 SE +/- 0.88, N = 3 SE +/- 21.79, N = 12 SE +/- 0.33, N = 3 SE +/- 0.58, N = 3 SE +/- 0.58, N = 3 SE +/- 0.33, N = 3 SE +/- 0.58, N = 3 SE +/- 11.06, N = 3 SE +/- 1.20, N = 3 SE +/- 7.84, N = 3 SE +/- 4.62, N = 3 531 561 661 759 894 954 961 1037 1049 1180 1342 1441 1497 1553 1657 1792 1988 2043 2090 2240 2318 3052 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 1.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.85 / Avg: 112.81 / Max: 142.09 Min: 18.36 / Avg: 131.28 / Max: 167.17 Min: 43.68 / Avg: 147 / Max: 223.13 Min: 43.56 / Avg: 188.73 / Max: 285.39 Min: 52 / Avg: 225.09 / Max: 290.77 Min: 22.93 / Avg: 248.39 / Max: 335.61 Min: 29.72 / Avg: 259.52 / Max: 331.33 Min: 35.06 / Avg: 288.94 / Max: 377.45 Min: 60.62 / Avg: 292.35 / Max: 377.02 Min: 75.94 / Avg: 303.61 / Max: 375.87 Min: 74.38 / Avg: 304.87 / Max: 378.48 Min: 56.57 / Avg: 313.79 / Max: 485.54 Min: 34.89 / Avg: 319.28 / Max: 410.78 Min: 108 / Avg: 365.31 / Max: 571.25 Min: 104.51 / Avg: 436.59 / Max: 581.52 Min: 41.9 / Avg: 518.09 / Max: 690.69 Min: 57.88 / Avg: 533.99 / Max: 677.83 Min: 81.36 / Avg: 571.36 / Max: 749.47 Min: 171.28 / Avg: 584.71 / Max: 751.6 Min: 198.09 / Avg: 602.84 / Max: 752.6 Min: 203.09 / Avg: 603.55 / Max: 753.72 Min: 62.5 / Avg: 640.62 / Max: 822.3
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 1 - Model: ResNet-50 Xeon Platinum 8380 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9654 EPYC 9684X Xeon Platinum 8592+ EPYC 9754 Xeon Max 9468 2P Xeon 6780E Xeon 6766E Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 4 8 12 16 20 SE +/- 0.08, N = 5 SE +/- 0.09, N = 4 SE +/- 0.08, N = 15 SE +/- 0.08, N = 15 SE +/- 0.05, N = 4 SE +/- 0.10, N = 4 SE +/- 0.13, N = 15 SE +/- 0.01, N = 4 SE +/- 0.09, N = 15 SE +/- 0.05, N = 9 SE +/- 0.05, N = 15 SE +/- 0.07, N = 3 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 SE +/- 0.06, N = 15 SE +/- 0.04, N = 15 SE +/- 0.01, N = 3 SE +/- 0.03, N = 15 SE +/- 0.06, N = 15 SE +/- 0.03, N = 15 SE +/- 0.02, N = 3 SE +/- 0.01, N = 15 17.82 11.29 10.79 10.70 10.53 9.99 9.13 9.01 8.96 6.43 6.19 5.11 5.02 4.67 4.39 4.37 3.92 3.85 3.66 2.47 1.99 1.76 images/sec Per Watt
OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 1 - Model: ResNet-50 EPYC 8534PN EPYC 8534P Xeon Platinum 8380 Xeon 6766E Xeon 6780E EPYC 9654 Xeon Platinum 8490H Xeon Max 9468 EPYC 9684X EPYC 9754 Xeon Max 9480 Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P Xeon Max 9480 2P 0.0315 0.063 0.0945 0.126 0.1575 0.140 0.132 0.130 0.062 0.060 0.057 0.056 0.053 0.050 0.048 0.045 0.042 0.032 0.011 0.011 0.010 0.010 0.010 0.009 0.009 0.009 0.009 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.16.1 CPU Power Consumption Monitor Xeon 6766E EPYC 8534PN EPYC 8534P Xeon 6780E EPYC 9754 Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9654 EPYC 9684X Xeon Platinum 8490H Xeon 6766E 2P Xeon 6780E 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 100 200 300 400 500 Min: 43.57 / Avg: 74.94 / Max: 90.44 Min: 15.23 / Avg: 76.32 / Max: 109.67 Min: 16.58 / Avg: 81.51 / Max: 118.9 Min: 43.51 / Avg: 83.13 / Max: 101.84 Min: 20.21 / Avg: 128.28 / Max: 166.19 Min: 49.72 / Avg: 137.42 / Max: 216.79 Min: 26.03 / Avg: 152.5 / Max: 203.99 Min: 29.81 / Avg: 157.26 / Max: 218.19 Min: 32.53 / Avg: 179.99 / Max: 250.99 Min: 59.45 / Avg: 188.55 / Max: 248.97 Min: 103.83 / Avg: 203.64 / Max: 238.14 Min: 104.65 / Avg: 212.41 / Max: 239.17 Min: 74.86 / Avg: 214.64 / Max: 277.01 Min: 74.4 / Avg: 220.35 / Max: 283.28 Min: 98.96 / Avg: 285.87 / Max: 395.19 Min: 39.45 / Avg: 288.27 / Max: 343.12 Min: 77.73 / Avg: 351.48 / Max: 431.04 Min: 55.24 / Avg: 357.87 / Max: 424.3 Min: 64.93 / Avg: 389.7 / Max: 463.09 Min: 139.48 / Avg: 455.97 / Max: 550.2 Min: 196.38 / Avg: 479.87 / Max: 557.8 Min: 193.89 / Avg: 495.88 / Max: 584.14
Result
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 64 - Model: ResNet-50 Xeon Platinum 8490H Xeon Platinum 8490H 2P EPYC 9654 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Max 9480 2P EPYC 9754 EPYC 9684X Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8380 EPYC 8534P EPYC 9684X 2P EPYC 9654 2P EPYC 8534PN Xeon 6780E Xeon 6766E Xeon 6780E 2P EPYC 9754 2P Xeon 6766E 2P 30 60 90 120 150 SE +/- 0.04, N = 3 SE +/- 1.12, N = 3 SE +/- 0.64, N = 3 SE +/- 0.39, N = 3 SE +/- 1.42, N = 3 SE +/- 1.28, N = 3 SE +/- 0.60, N = 3 SE +/- 0.08, N = 3 SE +/- 0.13, N = 3 SE +/- 0.93, N = 3 SE +/- 0.70, N = 3 SE +/- 0.50, N = 3 SE +/- 0.36, N = 3 SE +/- 0.05, N = 3 SE +/- 1.04, N = 4 SE +/- 0.91, N = 5 SE +/- 0.11, N = 3 SE +/- 0.14, N = 3 SE +/- 0.04, N = 3 SE +/- 0.76, N = 3 SE +/- 0.88, N = 3 SE +/- 0.56, N = 3 120.93 115.47 114.63 111.71 108.32 107.30 106.14 104.52 104.23 103.56 94.66 91.51 90.72 87.31 86.08 82.94 80.86 76.37 71.41 68.52 67.08 61.48 images/sec Per Watt
OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 64 - Model: ResNet-50 EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon 6766E Xeon Platinum 8490H Xeon Platinum 8592+ Xeon 6780E Xeon Platinum 8380 EPYC 9684X Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9468 2P EPYC 9754 2P Xeon Max 9480 2P EPYC 9654 2P EPYC 9684X 2P 0.1472 0.2944 0.4416 0.5888 0.736 0.654 0.610 0.477 0.448 0.423 0.398 0.382 0.378 0.362 0.354 0.330 0.326 0.236 0.202 0.202 0.188 0.186 0.186 0.179 0.178 0.170 0.159 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.16.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8380 EPYC 9654 EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 120 240 360 480 600 Min: 15.94 / Avg: 123.62 / Max: 139.7 Min: 17.24 / Avg: 143.22 / Max: 160.51 Min: 43.68 / Avg: 168.94 / Max: 189.35 Min: 43.53 / Avg: 201.8 / Max: 232.09 Min: 19.21 / Avg: 222.65 / Max: 256.54 Min: 35.02 / Avg: 239.37 / Max: 274.93 Min: 51.52 / Avg: 250.55 / Max: 277.16 Min: 28.64 / Avg: 256.02 / Max: 290.14 Min: 33.94 / Avg: 295.36 / Max: 331.29 Min: 57.29 / Avg: 303.56 / Max: 335.19 Min: 73.44 / Avg: 315.74 / Max: 344.17 Min: 75.88 / Avg: 317.98 / Max: 346.3 Min: 104.2 / Avg: 330.36 / Max: 365.9 Min: 53.44 / Avg: 364.61 / Max: 411.61 Min: 40.87 / Avg: 375.57 / Max: 410.97 Min: 104.27 / Avg: 458.39 / Max: 515.43 Min: 81.35 / Avg: 469.58 / Max: 533.37 Min: 53.67 / Avg: 487.46 / Max: 539.03 Min: 66.65 / Avg: 542.07 / Max: 602.63 Min: 158.09 / Avg: 571.99 / Max: 633.85 Min: 206.91 / Avg: 601.85 / Max: 661.69 Min: 193.17 / Avg: 603.74 / Max: 661.05
OpenVINO This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Machine Translation EN To DE FP16 - Device: CPU Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9480 2P EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8490H Xeon Platinum 8380 2P EPYC 9684X EPYC 9654 EPYC 8534P EPYC 9754 EPYC 8534PN Xeon Platinum 8380 Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 60 120 180 240 300 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.12, N = 3 SE +/- 0.07, N = 3 SE +/- 0.16, N = 3 SE +/- 0.34, N = 3 SE +/- 0.05, N = 3 SE +/- 0.01, N = 3 SE +/- 0.44, N = 3 SE +/- 0.59, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.60, N = 11 SE +/- 0.77, N = 7 SE +/- 0.13, N = 3 SE +/- 1.05, N = 5 SE +/- 0.05, N = 3 SE +/- 0.39, N = 3 SE +/- 0.06, N = 3 SE +/- 0.48, N = 3 SE +/- 3.14, N = 3 SE +/- 2.84, N = 15 26.85 28.17 28.73 29.98 37.58 38.50 41.57 43.17 48.43 56.73 62.46 70.95 83.11 91.16 100.67 102.84 109.00 127.89 240.38 259.66 271.14 294.69 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2024.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 Xeon Platinum 8490H EPYC 9654 Xeon Platinum 8592+ Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.99 / Avg: 135.1 / Max: 150.13 Min: 18.55 / Avg: 159.63 / Max: 174.23 Min: 43.66 / Avg: 230.17 / Max: 253.31 Min: 50.83 / Avg: 254.14 / Max: 284.65 Min: 43.55 / Avg: 296.98 / Max: 325.95 Min: 20.94 / Avg: 299.63 / Max: 344.93 Min: 59.14 / Avg: 321.95 / Max: 371.05 Min: 29.13 / Avg: 323.9 / Max: 362.47 Min: 34.43 / Avg: 325.81 / Max: 362.26 Min: 74.79 / Avg: 330.98 / Max: 373.36 Min: 70.61 / Avg: 331.05 / Max: 366.72 Min: 31.03 / Avg: 374.75 / Max: 410.94 Min: 106.93 / Avg: 441.51 / Max: 532 Min: 102.32 / Avg: 507.36 / Max: 566.08 Min: 108.68 / Avg: 553.41 / Max: 653.25 Min: 39.45 / Avg: 607.43 / Max: 679.21 Min: 52.92 / Avg: 645.83 / Max: 711.68 Min: 86.44 / Avg: 651.59 / Max: 748.25 Min: 167.2 / Avg: 656.03 / Max: 736.45 Min: 203.57 / Avg: 662.79 / Max: 751.8 Min: 196.83 / Avg: 663.81 / Max: 748.85 Min: 69.87 / Avg: 750.94 / Max: 819.6
OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Machine Translation EN To DE FP16 - Device: CPU Xeon Platinum 8592+ 2P EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P EPYC 9754 Xeon Platinum 8592+ EPYC 9684X EPYC 9654 Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8380 Xeon 6780E Xeon 6766E 200 400 600 800 1000 SE +/- 0.85, N = 3 SE +/- 11.78, N = 3 SE +/- 0.37, N = 3 SE +/- 8.96, N = 3 SE +/- 1.13, N = 3 SE +/- 1.99, N = 3 SE +/- 3.48, N = 3 SE +/- 6.34, N = 5 SE +/- 0.54, N = 3 SE +/- 4.29, N = 11 SE +/- 4.59, N = 7 SE +/- 0.17, N = 3 SE +/- 4.09, N = 3 SE +/- 1.21, N = 3 SE +/- 0.39, N = 3 SE +/- 0.13, N = 3 SE +/- 0.17, N = 3 SE +/- 3.00, N = 3 SE +/- 2.30, N = 15 SE +/- 0.45, N = 3 SE +/- 0.03, N = 3 SE +/- 0.25, N = 3 1128.78 1127.05 1110.49 990.16 889.18 795.63 793.64 621.42 592.20 576.98 525.88 477.77 467.08 415.96 317.38 293.07 281.36 264.66 243.76 156.13 149.63 138.53 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Parallel Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Max 9480 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P EPYC 9654 EPYC 9684X Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P EPYC 9754 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 8 16 24 32 40 SE +/- 0.01698, N = 3 SE +/- 0.01736, N = 3 SE +/- 0.00544, N = 3 SE +/- 0.02630, N = 3 SE +/- 0.03023, N = 3 SE +/- 0.03641, N = 15 SE +/- 0.03633, N = 15 SE +/- 0.03509, N = 3 SE +/- 0.06505, N = 4 SE +/- 0.04669, N = 3 SE +/- 0.01680, N = 3 SE +/- 0.02968, N = 3 SE +/- 0.06875, N = 3 SE +/- 0.03080, N = 3 SE +/- 0.02258, N = 3 SE +/- 0.05309, N = 3 SE +/- 0.05293, N = 3 SE +/- 0.08457, N = 3 SE +/- 0.07856, N = 3 SE +/- 0.08701, N = 3 SE +/- 0.32730, N = 6 SE +/- 0.34399, N = 3 3.13366 3.17360 3.19298 3.23780 3.32030 4.82506 4.92653 5.18070 5.25321 5.59491 6.00525 6.39031 6.49693 6.51744 6.52367 9.95245 9.96245 11.31480 15.58560 17.90000 35.22660 36.40090 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 EPYC 9654 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P 100 200 300 400 500 Min: 16.2 / Avg: 115 / Max: 124.26 Min: 18.11 / Avg: 126.52 / Max: 135.68 Min: 43.72 / Avg: 144.55 / Max: 152.6 Min: 43.63 / Avg: 170.55 / Max: 180.89 Min: 21.2 / Avg: 199.22 / Max: 213.69 Min: 53.36 / Avg: 238.84 / Max: 254.78 Min: 28.74 / Avg: 240.91 / Max: 259.49 Min: 36.05 / Avg: 285.19 / Max: 306.51 Min: 104.68 / Avg: 302.86 / Max: 321.76 Min: 108.09 / Avg: 347.86 / Max: 369.84 Min: 42.14 / Avg: 407.23 / Max: 436.3 Min: 101.46 / Avg: 445.87 / Max: 473.79 Min: 51.2 / Avg: 478.17 / Max: 512.01 Min: 64.35 / Avg: 547.96 / Max: 588.36
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Parallel Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Max 9480 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P EPYC 9654 EPYC 9684X Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P EPYC 9754 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 70 140 210 280 350 SE +/- 1.74, N = 3 SE +/- 1.73, N = 3 SE +/- 0.53, N = 3 SE +/- 2.50, N = 3 SE +/- 2.72, N = 3 SE +/- 1.54, N = 15 SE +/- 1.50, N = 15 SE +/- 1.30, N = 3 SE +/- 2.31, N = 4 SE +/- 1.50, N = 3 SE +/- 0.47, N = 3 SE +/- 0.73, N = 3 SE +/- 1.63, N = 3 SE +/- 0.73, N = 3 SE +/- 0.53, N = 3 SE +/- 0.54, N = 3 SE +/- 0.54, N = 3 SE +/- 0.66, N = 3 SE +/- 0.32, N = 3 SE +/- 0.27, N = 3 SE +/- 0.27, N = 6 SE +/- 0.26, N = 3 319.04 315.03 313.10 308.80 301.14 207.34 203.06 192.99 190.38 178.70 166.49 156.46 153.92 153.41 153.25 100.46 100.36 88.37 64.16 55.86 28.40 27.47 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Parallel Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Max 9480 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P EPYC 9654 EPYC 9684X Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P EPYC 9754 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 8 16 24 32 40 SE +/- 0.01698, N = 3 SE +/- 0.01736, N = 3 SE +/- 0.00544, N = 3 SE +/- 0.02630, N = 3 SE +/- 0.03023, N = 3 SE +/- 0.03641, N = 15 SE +/- 0.03633, N = 15 SE +/- 0.03509, N = 3 SE +/- 0.06505, N = 4 SE +/- 0.04669, N = 3 SE +/- 0.01680, N = 3 SE +/- 0.02968, N = 3 SE +/- 0.06875, N = 3 SE +/- 0.03080, N = 3 SE +/- 0.02258, N = 3 SE +/- 0.05309, N = 3 SE +/- 0.05293, N = 3 SE +/- 0.08457, N = 3 SE +/- 0.07856, N = 3 SE +/- 0.08701, N = 3 SE +/- 0.32730, N = 6 SE +/- 0.34399, N = 3 3.13366 3.17360 3.19298 3.23780 3.32030 4.82506 4.92653 5.18070 5.25321 5.59491 6.00525 6.39031 6.49693 6.51744 6.52367 9.95245 9.96245 11.31480 15.58560 17.90000 35.22660 36.40090 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OpenVINO This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Noise Suppression Poconet-Like FP16 - Device: CPU Xeon Platinum 8592+ Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8592+ 2P EPYC 9684X Xeon Platinum 8490H Xeon Platinum 8490H 2P EPYC 9654 Xeon Max 9468 2P EPYC 8534P EPYC 8534PN Xeon Max 9480 2P EPYC 9684X 2P EPYC 9754 EPYC 9654 2P Xeon Platinum 8380 Xeon Platinum 8380 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 70 140 210 280 350 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.05, N = 3 SE +/- 0.16, N = 14 SE +/- 0.12, N = 3 SE +/- 0.00, N = 3 SE +/- 0.02, N = 3 SE +/- 0.09, N = 3 SE +/- 0.04, N = 3 SE +/- 0.45, N = 12 SE +/- 0.08, N = 3 SE +/- 0.03, N = 3 SE +/- 0.07, N = 3 SE +/- 0.10, N = 3 SE +/- 0.65, N = 3 SE +/- 0.66, N = 3 SE +/- 0.86, N = 3 SE +/- 4.48, N = 3 9.40 11.37 12.72 13.15 13.52 13.95 15.96 16.32 17.15 17.30 18.49 18.72 22.20 22.67 23.12 25.91 26.33 28.10 75.22 82.12 293.36 340.26 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2024.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9754 Xeon 6766E 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6780E 2P EPYC 9684X EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.49 / Avg: 137.08 / Max: 149.27 Min: 17.46 / Avg: 161.25 / Max: 175.66 Min: 43.71 / Avg: 207.06 / Max: 230.09 Min: 43.5 / Avg: 248.78 / Max: 279.12 Min: 51.5 / Avg: 253.93 / Max: 289.36 Min: 20.78 / Avg: 292.63 / Max: 342.23 Min: 105.02 / Avg: 307.84 / Max: 385.97 Min: 29.75 / Avg: 315.18 / Max: 357.38 Min: 34.54 / Avg: 324.07 / Max: 405.42 Min: 60.51 / Avg: 327.19 / Max: 379.48 Min: 73.68 / Avg: 329.15 / Max: 372.33 Min: 85.86 / Avg: 330.63 / Max: 384.53 Min: 108.4 / Avg: 354.62 / Max: 434.65 Min: 34.79 / Avg: 374.52 / Max: 409.97 Min: 41.91 / Avg: 488.86 / Max: 570.14 Min: 103.92 / Avg: 501.31 / Max: 570.17 Min: 61.14 / Avg: 581.56 / Max: 658.32 Min: 43.35 / Avg: 637.42 / Max: 766.05 Min: 171.01 / Avg: 649.04 / Max: 743.64 Min: 204.87 / Avg: 656.57 / Max: 725.7 Min: 205.34 / Avg: 657.37 / Max: 743.06 Min: 67.53 / Avg: 716.78 / Max: 816.61
OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Noise Suppression Poconet-Like FP16 - Device: CPU Xeon Platinum 8592+ 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8490H 2P EPYC 9684X Xeon Platinum 8592+ EPYC 9754 2P Xeon Max 9480 2P EPYC 9654 Xeon Max 9468 2P EPYC 9754 Xeon Max 9480 Xeon Platinum 8490H Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon 6780E Xeon 6766E Xeon Platinum 8380 Xeon 6780E 2P Xeon 6766E 2P 2K 4K 6K 8K 10K SE +/- 24.76, N = 3 SE +/- 59.51, N = 3 SE +/- 77.59, N = 3 SE +/- 13.31, N = 3 SE +/- 1.84, N = 3 SE +/- 10.14, N = 3 SE +/- 27.90, N = 3 SE +/- 28.98, N = 3 SE +/- 67.22, N = 14 SE +/- 37.79, N = 3 SE +/- 139.41, N = 12 SE +/- 7.60, N = 3 SE +/- 1.87, N = 3 SE +/- 15.13, N = 3 SE +/- 0.64, N = 3 SE +/- 3.67, N = 3 SE +/- 6.96, N = 3 SE +/- 16.46, N = 3 SE +/- 13.67, N = 3 SE +/- 1.70, N = 3 SE +/- 2.88, N = 3 SE +/- 11.12, N = 3 9360.89 8231.72 7810.65 7356.63 6963.31 6757.26 6513.36 5933.65 5697.51 5565.23 5310.07 4386.48 4209.10 4208.49 3636.33 3405.94 2963.82 1906.99 1747.50 1525.57 978.46 843.92 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
libavif avifenc This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 1.0 Encoder Speed: 0 Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H Xeon Max 9480 EPYC 9754 EPYC 9754 2P EPYC 8534P Xeon Max 9468 EPYC 8534PN Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 30 60 90 120 150 SE +/- 0.03, N = 3 SE +/- 0.12, N = 3 SE +/- 0.01, N = 3 SE +/- 0.11, N = 3 SE +/- 0.01, N = 3 SE +/- 0.08, N = 3 SE +/- 0.05, N = 3 SE +/- 0.32, N = 3 SE +/- 0.29, N = 3 SE +/- 0.32, N = 3 SE +/- 0.22, N = 3 SE +/- 0.13, N = 3 SE +/- 0.06, N = 3 SE +/- 0.19, N = 3 SE +/- 0.15, N = 3 SE +/- 0.18, N = 3 SE +/- 0.18, N = 3 SE +/- 0.52, N = 3 SE +/- 0.12, N = 3 SE +/- 0.18, N = 3 SE +/- 0.03, N = 3 SE +/- 0.22, N = 3 60.53 61.63 61.99 62.89 63.40 64.48 67.38 68.44 69.65 69.93 73.06 73.27 74.69 75.86 76.92 78.24 88.07 100.46 126.97 129.91 140.86 143.80 1. (CXX) g++ options: -O3 -fPIC -lm CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better libavif avifenc 1.0 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 43.56 / Avg: 65.01 / Max: 148.96 Min: 43.39 / Avg: 72.37 / Max: 182.54 Min: 16.02 / Avg: 77.61 / Max: 131.52 Min: 17.01 / Avg: 85.23 / Max: 156.74 Min: 19.55 / Avg: 121.7 / Max: 248.44 Min: 104.01 / Avg: 131.11 / Max: 231.48 Min: 105.44 / Avg: 139.04 / Max: 263.97 Min: 29.48 / Avg: 156.52 / Max: 318.36 Min: 49.17 / Avg: 165.96 / Max: 290.88 Min: 30.24 / Avg: 180.87 / Max: 389.03 Min: 34.78 / Avg: 197.77 / Max: 373.2 Min: 61.27 / Avg: 218.41 / Max: 375.95 Min: 39.64 / Avg: 234.99 / Max: 370.06 Min: 69.59 / Avg: 253.47 / Max: 376.81 Min: 75.95 / Avg: 254.61 / Max: 372.81 Min: 96.24 / Avg: 285.14 / Max: 552.25 Min: 58.14 / Avg: 295.3 / Max: 499.26 Min: 62.98 / Avg: 324.3 / Max: 548.71 Min: 77.6 / Avg: 391.09 / Max: 728.04 Min: 152.89 / Avg: 425.04 / Max: 708.08 Min: 208.88 / Avg: 485.4 / Max: 727.55 Min: 205.63 / Avg: 494.12 / Max: 704.89
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: T5 Encoder - Device: CPU - Executor: Parallel Xeon Platinum 8592+ EPYC 9684X EPYC 9654 Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H EPYC 9754 Xeon Platinum 8380 EPYC 9684X 2P EPYC 9654 2P EPYC 8534P EPYC 8534PN Xeon Max 9480 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 3 6 9 12 15 SE +/- 0.01801, N = 3 SE +/- 0.01812, N = 3 SE +/- 0.00389, N = 3 SE +/- 0.01381, N = 3 SE +/- 0.01443, N = 3 SE +/- 0.03338, N = 15 SE +/- 0.00807, N = 3 SE +/- 0.02990, N = 3 SE +/- 0.02857, N = 3 SE +/- 0.00852, N = 3 SE +/- 0.03309, N = 3 SE +/- 0.03399, N = 3 SE +/- 0.02993, N = 3 SE +/- 0.02633, N = 15 SE +/- 0.01019, N = 3 SE +/- 0.01782, N = 3 SE +/- 0.04072, N = 3 SE +/- 0.04801, N = 3 SE +/- 0.01135, N = 3 SE +/- 0.01143, N = 3 SE +/- 0.14331, N = 3 SE +/- 0.12596, N = 3 2.15980 2.23922 2.66711 2.81739 2.88734 2.97094 3.06500 3.12904 3.54242 3.61792 3.68001 3.91006 3.92455 3.95819 3.98735 4.00675 4.05270 4.05881 4.06272 4.20355 10.58830 10.80660 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8380 EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P 60 120 180 240 300 Min: 15.18 / Avg: 80.15 / Max: 85.73 Min: 17.28 / Avg: 82.81 / Max: 89.26 Min: 43.59 / Avg: 101.11 / Max: 106.64 Min: 43.42 / Avg: 105.34 / Max: 111.27 Min: 21.09 / Avg: 108.61 / Max: 116.43 Min: 29.55 / Avg: 141.28 / Max: 150.78 Min: 33.5 / Avg: 163.65 / Max: 174.8 Min: 103.76 / Avg: 180.19 / Max: 189.22 Min: 105.65 / Avg: 182.76 / Max: 192.69 Min: 52.52 / Avg: 218.31 / Max: 234.43 Min: 40.04 / Avg: 220.47 / Max: 233.47 Min: 57.25 / Avg: 267.87 / Max: 282.78 Min: 64.84 / Avg: 297.41 / Max: 317.19 Min: 99.27 / Avg: 314.65 / Max: 333.57
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: T5 Encoder - Device: CPU - Executor: Parallel Xeon Platinum 8592+ EPYC 9684X EPYC 9654 Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H EPYC 9754 Xeon Platinum 8380 EPYC 9684X 2P EPYC 9654 2P EPYC 8534P EPYC 8534PN Xeon Max 9480 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 100 200 300 400 500 SE +/- 3.84, N = 3 SE +/- 3.58, N = 3 SE +/- 0.55, N = 3 SE +/- 1.73, N = 3 SE +/- 1.74, N = 3 SE +/- 3.48, N = 15 SE +/- 0.86, N = 3 SE +/- 3.02, N = 3 SE +/- 2.25, N = 3 SE +/- 0.65, N = 3 SE +/- 2.43, N = 3 SE +/- 2.24, N = 3 SE +/- 1.96, N = 3 SE +/- 1.67, N = 15 SE +/- 0.64, N = 3 SE +/- 1.11, N = 3 SE +/- 2.44, N = 3 SE +/- 2.94, N = 3 SE +/- 0.69, N = 3 SE +/- 0.65, N = 3 SE +/- 1.28, N = 3 SE +/- 1.08, N = 3 462.86 446.30 374.75 354.75 346.21 337.00 326.14 319.45 282.25 276.21 271.62 255.67 254.73 252.67 250.63 249.48 246.68 246.34 246.04 237.80 94.42 92.51 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: T5 Encoder - Device: CPU - Executor: Parallel Xeon Platinum 8592+ EPYC 9684X EPYC 9654 Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H EPYC 9754 Xeon Platinum 8380 EPYC 9684X 2P EPYC 9654 2P EPYC 8534P EPYC 8534PN Xeon Max 9480 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 3 6 9 12 15 SE +/- 0.01801, N = 3 SE +/- 0.01812, N = 3 SE +/- 0.00389, N = 3 SE +/- 0.01381, N = 3 SE +/- 0.01443, N = 3 SE +/- 0.03338, N = 15 SE +/- 0.00807, N = 3 SE +/- 0.02990, N = 3 SE +/- 0.02857, N = 3 SE +/- 0.00852, N = 3 SE +/- 0.03309, N = 3 SE +/- 0.03399, N = 3 SE +/- 0.02993, N = 3 SE +/- 0.02633, N = 15 SE +/- 0.01019, N = 3 SE +/- 0.01782, N = 3 SE +/- 0.04072, N = 3 SE +/- 0.04801, N = 3 SE +/- 0.01135, N = 3 SE +/- 0.01143, N = 3 SE +/- 0.14331, N = 3 SE +/- 0.12596, N = 3 2.15980 2.23922 2.66711 2.81739 2.88734 2.97094 3.06500 3.12904 3.54242 3.61792 3.68001 3.91006 3.92455 3.95819 3.98735 4.00675 4.05270 4.05881 4.06272 4.20355 10.58830 10.80660 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
ASKAP ASKAP is a set of benchmarks from the Australian SKA Pathfinder. The principal ASKAP benchmarks are the Hogbom Clean Benchmark (tHogbomClean) and Convolutional Resamping Benchmark (tConvolve) as well as some previous ASKAP benchmarks being included as well for OpenCL and CUDA execution of tConvolve. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Mpix/sec, More Is Better ASKAP 1.0 Test: tConvolve MPI - Gridding EPYC 9684X 2P EPYC 9654 2P EPYC 9684X Xeon Platinum 8592+ 2P EPYC 9754 2P EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ EPYC 9754 Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon Platinum 8380 30K 60K 90K 120K 150K SE +/- 19579.11, N = 3 SE +/- 2386.93, N = 4 SE +/- 581.17, N = 3 SE +/- 1080.03, N = 3 SE +/- 1180.69, N = 14 SE +/- 1853.72, N = 3 SE +/- 802.88, N = 3 SE +/- 194.30, N = 3 SE +/- 412.68, N = 4 SE +/- 219.20, N = 3 SE +/- 685.49, N = 12 SE +/- 0.00, N = 3 SE +/- 99.83, N = 3 SE +/- 76.67, N = 3 SE +/- 86.77, N = 3 SE +/- 0.00, N = 3 SE +/- 46.25, N = 3 SE +/- 47.29, N = 3 121976.90 86757.70 74389.00 72004.00 54066.00 46233.70 45195.10 42939.20 41773.00 37265.90 28854.30 23250.20 21766.50 21314.20 16573.20 16399.70 14497.30 9623.77 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp Mpix/sec Per Watt
OpenBenchmarking.org Mpix/sec Per Watt, More Is Better ASKAP 1.0 Test: tConvolve MPI - Gridding EPYC 9684X EPYC 9684X 2P EPYC 9654 EPYC 9654 2P EPYC 8534PN EPYC 9754 EPYC 8534P Xeon Platinum 8592+ 2P Xeon Platinum 8592+ EPYC 9754 2P Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 50 100 150 200 250 247.30 199.37 188.14 167.00 142.12 134.70 130.96 128.04 125.14 109.74 75.14 70.87 69.96 67.85 67.77 66.72 38.76 29.22 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ASKAP 1.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9684X Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 16.5 / Avg: 115.4 / Max: 136.82 Min: 19.03 / Avg: 126.56 / Max: 161.28 Min: 20.61 / Avg: 214.21 / Max: 327.89 Min: 29.47 / Avg: 245.74 / Max: 336.85 Min: 52.74 / Avg: 248.29 / Max: 286.59 Min: 24.59 / Avg: 297.8 / Max: 352.67 Min: 31.29 / Avg: 300.8 / Max: 409.34 Min: 73.2 / Avg: 309.44 / Max: 360.17 Min: 74.39 / Avg: 311.12 / Max: 354.1 Min: 57.81 / Avg: 314.14 / Max: 357.36 Min: 37.58 / Avg: 492.69 / Max: 670.56 Min: 97.93 / Avg: 496.22 / Max: 563.93 Min: 57.21 / Avg: 519.5 / Max: 681.6 Min: 74.42 / Avg: 562.35 / Max: 727.39 Min: 64.2 / Avg: 611.81 / Max: 819.12 Min: 163.2 / Avg: 626.07 / Max: 724.87 Min: 196.26 / Avg: 633.57 / Max: 720.17 Min: 196.89 / Avg: 637.73 / Max: 729.76
OpenBenchmarking.org Mpix/sec, More Is Better ASKAP 1.0 Test: tConvolve MPI - Degridding EPYC 9684X 2P EPYC 9654 2P EPYC 9684X Xeon Platinum 8592+ 2P EPYC 9754 2P EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ EPYC 9754 Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon Platinum 8380 30K 60K 90K 120K 150K SE +/- 0.00, N = 3 SE +/- 882.91, N = 4 SE +/- 0.00, N = 3 SE +/- 480.95, N = 3 SE +/- 1077.92, N = 14 SE +/- 189.10, N = 3 SE +/- 0.00, N = 3 SE +/- 143.60, N = 3 SE +/- 378.22, N = 4 SE +/- 153.57, N = 3 SE +/- 598.81, N = 12 SE +/- 76.47, N = 3 SE +/- 0.00, N = 3 SE +/- 93.91, N = 3 SE +/- 44.43, N = 3 SE +/- 69.57, N = 3 SE +/- 62.13, N = 3 SE +/- 64.71, N = 3 123480.00 81025.80 64260.20 59139.20 52639.50 42172.40 40368.60 36900.50 34614.10 31177.20 28476.60 20561.40 18522.00 17891.60 16704.40 15903.40 12800.40 8429.13 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
Helsing Helsing is an open-source POSIX vampire number generator. This test profile measures the time it takes to generate vampire numbers between varying numbers of digits. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Helsing 1.0-beta Digit Range: 14 digit EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9754 Xeon Platinum 8490H 2P Xeon 6780E 2P EPYC 9684X Xeon Max 9480 2P Xeon 6766E 2P EPYC 9654 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6780E EPYC 8534P Xeon Max 9480 Xeon 6766E EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 40 80 120 160 200 SE +/- 0.32, N = 3 SE +/- 0.42, N = 3 SE +/- 0.38, N = 5 SE +/- 0.07, N = 3 SE +/- 0.02, N = 3 SE +/- 0.06, N = 3 SE +/- 0.01, N = 3 SE +/- 0.14, N = 3 SE +/- 0.42, N = 3 SE +/- 0.11, N = 3 SE +/- 0.10, N = 3 SE +/- 0.11, N = 3 SE +/- 0.09, N = 3 SE +/- 0.08, N = 3 SE +/- 0.26, N = 3 SE +/- 0.10, N = 3 SE +/- 0.27, N = 3 SE +/- 0.12, N = 3 SE +/- 0.12, N = 3 SE +/- 0.06, N = 3 SE +/- 0.53, N = 3 SE +/- 0.08, N = 3 26.78 30.69 33.68 48.09 48.85 50.83 53.02 56.91 58.25 58.96 62.17 65.03 82.96 89.07 93.35 103.86 105.68 107.24 115.46 120.56 123.83 165.79 1. (CC) gcc options: -O2 -pthread CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Helsing 1.0-beta CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 150 300 450 600 750 Min: 17.86 / Avg: 131.19 / Max: 137.88 Min: 18.11 / Avg: 166.23 / Max: 176.55 Min: 43.76 / Avg: 217.84 / Max: 231.94 Min: 52.7 / Avg: 264.22 / Max: 291.17 Min: 43.52 / Avg: 283.2 / Max: 302.38 Min: 21.32 / Avg: 315.34 / Max: 347.42 Min: 28.89 / Avg: 322.19 / Max: 348.84 Min: 32.85 / Avg: 334.81 / Max: 377.7 Min: 56.9 / Avg: 336.32 / Max: 376.98 Min: 87.46 / Avg: 338.48 / Max: 364.16 Min: 73.73 / Avg: 338.57 / Max: 364.52 Min: 34.66 / Avg: 378.16 / Max: 412.26 Min: 105.07 / Avg: 444.27 / Max: 495.95 Min: 102.73 / Avg: 517.9 / Max: 581.32 Min: 107.47 / Avg: 552.92 / Max: 634.15 Min: 41.44 / Avg: 591.29 / Max: 696.59 Min: 58.49 / Avg: 624.24 / Max: 716.4 Min: 74.95 / Avg: 643.2 / Max: 755.25 Min: 173.59 / Avg: 654.95 / Max: 756.78 Min: 203.6 / Avg: 662.24 / Max: 747.53 Min: 210.37 / Avg: 664.49 / Max: 742.91 Min: 67.93 / Avg: 711.96 / Max: 830.1
OSPRay Studio Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 1.0 Camera: 1 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPU EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9754 Xeon Max 9480 2P EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon 6780E 2P Xeon Max 9468 Xeon 6766E 2P EPYC 8534P EPYC 8534PN Xeon 6780E Xeon 6766E Xeon Platinum 8380 12K 24K 36K 48K 60K SE +/- 12.34, N = 3 SE +/- 17.16, N = 3 SE +/- 32.74, N = 3 SE +/- 7.13, N = 3 SE +/- 34.10, N = 3 SE +/- 33.33, N = 3 SE +/- 51.39, N = 3 SE +/- 39.30, N = 3 SE +/- 72.42, N = 3 SE +/- 44.33, N = 3 SE +/- 37.49, N = 3 SE +/- 53.03, N = 3 SE +/- 62.46, N = 3 SE +/- 83.47, N = 3 SE +/- 429.71, N = 4 SE +/- 28.06, N = 3 SE +/- 409.89, N = 15 SE +/- 250.73, N = 3 SE +/- 31.02, N = 3 SE +/- 371.33, N = 3 SE +/- 199.93, N = 3 SE +/- 148.46, N = 3 8413 8916 10428 12124 14138 15286 15357 16603 16645 18896 23125 30442 31380 33790 35081 37055 38228 38811 42513 46875 52083 54456 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 1.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9754 Xeon 6766E 2P Xeon Platinum 8490H EPYC 9654 Xeon 6780E 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.28 / Avg: 103.3 / Max: 141.73 Min: 17.86 / Avg: 118.24 / Max: 166.58 Min: 43.63 / Avg: 134.74 / Max: 219.04 Min: 43.51 / Avg: 165.73 / Max: 285.29 Min: 53.95 / Avg: 217.25 / Max: 291.65 Min: 21.92 / Avg: 245.02 / Max: 336.36 Min: 105.3 / Avg: 248.8 / Max: 487.82 Min: 61.9 / Avg: 259.12 / Max: 377.23 Min: 29.03 / Avg: 265.51 / Max: 331.71 Min: 107.79 / Avg: 277.62 / Max: 555.46 Min: 78.1 / Avg: 281.43 / Max: 376.32 Min: 74.46 / Avg: 283.63 / Max: 375.94 Min: 35.35 / Avg: 299.63 / Max: 376.51 Min: 35.54 / Avg: 320.39 / Max: 411.05 Min: 95.78 / Avg: 375.01 / Max: 583.47 Min: 61.04 / Avg: 548.32 / Max: 683.37 Min: 41.91 / Avg: 564.91 / Max: 692.23 Min: 78.22 / Avg: 573.74 / Max: 753.21 Min: 201.01 / Avg: 599.93 / Max: 754.22 Min: 210.55 / Avg: 606.6 / Max: 753.89 Min: 175.84 / Avg: 622.79 / Max: 756.66 Min: 66.57 / Avg: 696.59 / Max: 822.15
Result
OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 1.0 Camera: 3 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPU EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9754 EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon 6780E 2P Xeon Max 9468 Xeon 6766E 2P EPYC 8534P EPYC 8534PN Xeon 6780E Xeon 6766E Xeon Platinum 8380 14K 28K 42K 56K 70K SE +/- 16.90, N = 3 SE +/- 32.79, N = 3 SE +/- 27.79, N = 3 SE +/- 8.00, N = 3 SE +/- 29.96, N = 3 SE +/- 10.97, N = 3 SE +/- 21.05, N = 3 SE +/- 24.66, N = 3 SE +/- 25.12, N = 3 SE +/- 5.55, N = 3 SE +/- 26.00, N = 3 SE +/- 28.75, N = 3 SE +/- 61.20, N = 3 SE +/- 74.33, N = 3 SE +/- 301.00, N = 12 SE +/- 36.47, N = 3 SE +/- 51.90, N = 3 SE +/- 111.72, N = 3 SE +/- 197.99, N = 3 SE +/- 26.58, N = 3 SE +/- 500.31, N = 3 SE +/- 62.75, N = 3 10015 10514 12427 14358 16837 18111 18129 19535 19701 22305 31993 34866 36649 39238 39283 42837 43906 45377 48844 52501 57657 63035 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 1.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9754 Xeon 6766E 2P Xeon Platinum 8490H EPYC 9654 Xeon Max 9480 Xeon Max 9468 Xeon 6780E 2P EPYC 9684X Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.31 / Avg: 107.81 / Max: 142.82 Min: 17.33 / Avg: 121.97 / Max: 166.66 Min: 43.65 / Avg: 142.62 / Max: 223.55 Min: 43.5 / Avg: 175.23 / Max: 284.38 Min: 53.28 / Avg: 223.15 / Max: 291.09 Min: 26.59 / Avg: 252.77 / Max: 377.16 Min: 22.55 / Avg: 256.1 / Max: 337.04 Min: 105.47 / Avg: 265.24 / Max: 464.47 Min: 59.63 / Avg: 268.15 / Max: 377.52 Min: 27.99 / Avg: 273.01 / Max: 332.14 Min: 76.02 / Avg: 287.68 / Max: 377.28 Min: 68.84 / Avg: 291.26 / Max: 378.07 Min: 107.96 / Avg: 295.44 / Max: 567.47 Min: 35.75 / Avg: 329.8 / Max: 411.04 Min: 106.2 / Avg: 388.48 / Max: 582.9 Min: 42.24 / Avg: 558.27 / Max: 695.79 Min: 63.21 / Avg: 573.82 / Max: 678.02 Min: 175.58 / Avg: 588.27 / Max: 757.42 Min: 199.93 / Avg: 609.51 / Max: 752.77 Min: 210.27 / Avg: 614.16 / Max: 754.39 Min: 78.39 / Avg: 616.75 / Max: 753.89 Min: 63.78 / Avg: 623.33 / Max: 822.16
CloverLeaf CloverLeaf is a Lagrangian-Eulerian hydrodynamics benchmark. This test profile currently makes use of CloverLeaf's OpenMP version. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better CloverLeaf 1.3 Input: clover_bm64_short EPYC 9684X 2P Xeon Platinum 8490H 2P EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8380 2P Xeon 6766E Xeon 6780E Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 EPYC 9684X Xeon Max 9480 EPYC 9654 EPYC 9754 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 14 28 42 56 70 SE +/- 0.23, N = 3 SE +/- 0.24, N = 3 SE +/- 0.30, N = 3 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 SE +/- 1.11, N = 12 SE +/- 0.45, N = 3 SE +/- 0.31, N = 15 SE +/- 0.40, N = 4 SE +/- 0.31, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.15, N = 3 SE +/- 1.81, N = 15 SE +/- 0.17, N = 3 SE +/- 2.15, N = 15 SE +/- 2.07, N = 12 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 22.83 26.58 27.74 28.92 28.99 30.05 31.90 32.25 32.44 35.29 38.01 38.01 39.38 39.48 41.43 42.17 42.22 46.50 48.36 57.38 57.54 60.89 1. (F9X) gfortran options: -O3 -march=native -funroll-loops -fopenmp CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better CloverLeaf 1.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 17.1 / Avg: 122.03 / Max: 133.14 Min: 19.22 / Avg: 132.55 / Max: 146.42 Min: 43.65 / Avg: 168.7 / Max: 186.26 Min: 19.77 / Avg: 183.89 / Max: 245.13 Min: 43.45 / Avg: 184.14 / Max: 204.33 Min: 25.91 / Avg: 223.18 / Max: 288.87 Min: 53.35 / Avg: 254.93 / Max: 281.37 Min: 33.05 / Avg: 264.63 / Max: 353.22 Min: 32.08 / Avg: 299.62 / Max: 336.85 Min: 57.26 / Avg: 320.54 / Max: 366.56 Min: 74.72 / Avg: 323.64 / Max: 372.82 Min: 76.4 / Avg: 324.2 / Max: 367.03 Min: 104.9 / Avg: 364.67 / Max: 444.64 Min: 42.14 / Avg: 366.92 / Max: 461.67 Min: 107.36 / Avg: 389.09 / Max: 472.12 Min: 56.35 / Avg: 437.97 / Max: 551.54 Min: 102.86 / Avg: 479 / Max: 543.07 Min: 68.07 / Avg: 499.33 / Max: 638.11 Min: 75.65 / Avg: 552.24 / Max: 662.73 Min: 166.73 / Avg: 611.92 / Max: 712.51 Min: 195.63 / Avg: 623.99 / Max: 717.07 Min: 199.64 / Avg: 627.92 / Max: 713.69
OSPRay Studio Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OSPRay Studio 1.0 Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPU EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9754 Xeon Max 9480 2P EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon 6780E 2P Xeon Platinum 8592+ Xeon 6766E 2P Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 8534P Xeon 6780E EPYC 8534PN Xeon 6766E Xeon Platinum 8380 800 1600 2400 3200 4000 SE +/- 0.67, N = 3 SE +/- 1.45, N = 3 SE +/- 0.58, N = 3 SE +/- 0.00, N = 3 SE +/- 0.88, N = 3 SE +/- 0.88, N = 3 SE +/- 0.33, N = 3 SE +/- 1.53, N = 3 SE +/- 2.65, N = 3 SE +/- 1.00, N = 3 SE +/- 16.65, N = 5 SE +/- 1.45, N = 3 SE +/- 4.67, N = 3 SE +/- 2.73, N = 3 SE +/- 0.88, N = 3 SE +/- 1.20, N = 3 SE +/- 1.20, N = 3 SE +/- 1.00, N = 3 SE +/- 3.51, N = 3 SE +/- 3.06, N = 3 SE +/- 6.23, N = 3 SE +/- 3.79, N = 3 630 660 780 898 1063 1130 1143 1225 1237 1391 1631 1709 1832 1833 1977 2123 2357 2421 2472 2649 2731 3616 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OSPRay Studio 1.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.17 / Avg: 114.16 / Max: 142.74 Min: 18.61 / Avg: 131.5 / Max: 167.12 Min: 43.67 / Avg: 146.96 / Max: 222.56 Min: 43.6 / Avg: 190.51 / Max: 286.01 Min: 54.18 / Avg: 227.88 / Max: 291.87 Min: 22.69 / Avg: 249.28 / Max: 335.5 Min: 31.26 / Avg: 263.32 / Max: 331.82 Min: 26.01 / Avg: 290.77 / Max: 377.76 Min: 61.31 / Avg: 293.93 / Max: 377.6 Min: 74.64 / Avg: 306.2 / Max: 376.48 Min: 71.66 / Avg: 307.39 / Max: 377.12 Min: 105.7 / Avg: 310.43 / Max: 465.96 Min: 30.75 / Avg: 323.21 / Max: 411.04 Min: 108.07 / Avg: 363.45 / Max: 555.01 Min: 103.42 / Avg: 440.54 / Max: 582.66 Min: 42.39 / Avg: 515 / Max: 688.22 Min: 59.99 / Avg: 534.76 / Max: 676.92 Min: 79.09 / Avg: 574.5 / Max: 751.34 Min: 179.12 / Avg: 588.21 / Max: 753.37 Min: 204.32 / Avg: 605.57 / Max: 755.38 Min: 208.56 / Avg: 606.68 / Max: 753.95 Min: 67.19 / Avg: 642.74 / Max: 821.11
OpenVINO This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Road Segmentation ADAS FP16-INT8 - Device: CPU EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9684X EPYC 9654 Xeon Platinum 8380 EPYC 8534P EPYC 8534PN Xeon 6780E Xeon 6780E 2P EPYC 9754 Xeon 6766E Xeon 6766E 2P Xeon Platinum 8592+ Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Max 9480 Xeon Platinum 8490H 2P 15 30 45 60 75 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 SE +/- 0.01, N = 3 SE +/- 0.11, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.10, N = 3 SE +/- 0.04, N = 3 SE +/- 0.26, N = 15 SE +/- 0.11, N = 3 SE +/- 0.05, N = 3 SE +/- 0.02, N = 3 SE +/- 0.15, N = 3 SE +/- 0.14, N = 3 SE +/- 0.05, N = 3 SE +/- 0.28, N = 3 SE +/- 0.06, N = 3 SE +/- 0.21, N = 3 SE +/- 0.03, N = 3 11.95 14.30 15.84 16.84 21.63 27.46 31.05 31.99 34.76 35.30 37.17 38.21 38.87 41.50 55.07 55.74 55.99 56.71 57.13 62.11 62.53 65.02 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2024.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.95 / Avg: 130.31 / Max: 141.74 Min: 16.37 / Avg: 152.34 / Max: 164.44 Min: 43.66 / Avg: 218.36 / Max: 239.94 Min: 50.28 / Avg: 253.24 / Max: 288.06 Min: 21.36 / Avg: 269.44 / Max: 322.88 Min: 43.56 / Avg: 269.61 / Max: 298.84 Min: 30.63 / Avg: 302.4 / Max: 328.21 Min: 34.63 / Avg: 325.47 / Max: 376.24 Min: 59.15 / Avg: 327.9 / Max: 376.83 Min: 84.64 / Avg: 330.18 / Max: 380.81 Min: 74.41 / Avg: 330.18 / Max: 376.61 Min: 33.63 / Avg: 370.93 / Max: 400.44 Min: 106.85 / Avg: 458.61 / Max: 523.13 Min: 99.46 / Avg: 509.59 / Max: 578.94 Min: 109.15 / Avg: 564.61 / Max: 636.36 Min: 40.64 / Avg: 591.23 / Max: 648.1 Min: 60.61 / Avg: 599.17 / Max: 650.25 Min: 81.23 / Avg: 641.35 / Max: 755.31 Min: 168.6 / Avg: 652.87 / Max: 754.8 Min: 200.42 / Avg: 659.67 / Max: 754.38 Min: 201 / Avg: 659.95 / Max: 753.12 Min: 67.8 / Avg: 760.09 / Max: 819.06
OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Road Segmentation ADAS FP16-INT8 - Device: CPU EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9654 Xeon 6766E 2P Xeon Max 9468 2P EPYC 9754 Xeon Platinum 8380 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6780E EPYC 8534P Xeon 6766E EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 900 1800 2700 3600 4500 SE +/- 5.40, N = 3 SE +/- 5.33, N = 3 SE +/- 3.90, N = 3 SE +/- 11.53, N = 3 SE +/- 1.09, N = 3 SE +/- 2.23, N = 3 SE +/- 0.66, N = 3 SE +/- 1.74, N = 3 SE +/- 6.85, N = 3 SE +/- 2.08, N = 3 SE +/- 1.56, N = 3 SE +/- 11.41, N = 15 SE +/- 3.75, N = 3 SE +/- 0.47, N = 3 SE +/- 2.65, N = 3 SE +/- 2.82, N = 3 SE +/- 0.65, N = 3 SE +/- 2.69, N = 3 SE +/- 0.81, N = 3 SE +/- 2.94, N = 3 SE +/- 2.29, N = 3 SE +/- 0.62, N = 3 4032.66 4013.96 3353.73 2237.69 2212.41 1933.08 1843.61 1801.89 1743.49 1731.50 1691.79 1671.23 1186.24 1159.66 1068.43 1018.18 998.09 924.90 918.53 894.87 860.64 642.44 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: ArcFace ResNet-100 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 Xeon Max 9480 2P Xeon Platinum 8490H 2P EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 60 120 180 240 300 SE +/- 0.07, N = 3 SE +/- 0.04, N = 3 SE +/- 0.06, N = 3 SE +/- 0.12, N = 3 SE +/- 0.17, N = 3 SE +/- 0.37, N = 3 SE +/- 0.11, N = 3 SE +/- 0.28, N = 3 SE +/- 0.05, N = 3 SE +/- 0.04, N = 3 SE +/- 0.13, N = 3 SE +/- 0.11, N = 3 SE +/- 0.17, N = 3 SE +/- 0.35, N = 3 SE +/- 0.13, N = 3 SE +/- 0.22, N = 3 SE +/- 0.25, N = 3 SE +/- 0.45, N = 3 SE +/- 0.89, N = 3 SE +/- 0.84, N = 3 SE +/- 1.74, N = 15 SE +/- 3.87, N = 3 16.55 21.34 22.59 23.66 24.55 25.98 30.47 31.79 33.50 34.16 34.25 34.31 36.61 38.10 40.63 66.44 66.63 83.77 146.78 160.13 266.10 285.99 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P 100 200 300 400 500 Min: 16.96 / Avg: 119.46 / Max: 129.05 Min: 19.18 / Avg: 133.31 / Max: 143.64 Min: 43.73 / Avg: 146.66 / Max: 157.5 Min: 43.59 / Avg: 175.34 / Max: 190.81 Min: 21.46 / Avg: 198.94 / Max: 215.69 Min: 30.05 / Avg: 245.24 / Max: 265.47 Min: 51.87 / Avg: 247.61 / Max: 266.6 Min: 34.21 / Avg: 290.09 / Max: 313.77 Min: 104.82 / Avg: 306.33 / Max: 331.18 Min: 107.98 / Avg: 351.03 / Max: 380.24 Min: 35.54 / Avg: 404.89 / Max: 435.36 Min: 102.55 / Avg: 456.44 / Max: 489.76 Min: 55.89 / Avg: 483.5 / Max: 517.8 Min: 62.47 / Avg: 550.79 / Max: 593.02
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: ArcFace ResNet-100 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 Xeon Max 9480 2P Xeon Platinum 8490H 2P EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 14 28 42 56 70 SE +/- 0.27014, N = 3 SE +/- 0.09731, N = 3 SE +/- 0.10899, N = 3 SE +/- 0.20934, N = 3 SE +/- 0.27882, N = 3 SE +/- 0.55506, N = 3 SE +/- 0.11901, N = 3 SE +/- 0.27122, N = 3 SE +/- 0.04145, N = 3 SE +/- 0.03174, N = 3 SE +/- 0.11212, N = 3 SE +/- 0.09239, N = 3 SE +/- 0.12474, N = 3 SE +/- 0.23936, N = 3 SE +/- 0.07716, N = 3 SE +/- 0.05013, N = 3 SE +/- 0.05569, N = 3 SE +/- 0.06474, N = 3 SE +/- 0.04131, N = 3 SE +/- 0.03255, N = 3 SE +/- 0.02474, N = 15 SE +/- 0.04782, N = 3 60.40940 46.84910 44.27480 42.27110 40.72890 38.50250 32.81410 31.45890 29.84930 29.26900 29.19550 29.14870 27.31280 26.24750 24.60950 15.05160 15.00820 11.93820 6.81330 6.24521 3.76003 3.49785 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: ArcFace ResNet-100 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 Xeon Max 9480 2P Xeon Platinum 8490H 2P EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 60 120 180 240 300 SE +/- 0.07, N = 3 SE +/- 0.04, N = 3 SE +/- 0.06, N = 3 SE +/- 0.12, N = 3 SE +/- 0.17, N = 3 SE +/- 0.37, N = 3 SE +/- 0.11, N = 3 SE +/- 0.28, N = 3 SE +/- 0.05, N = 3 SE +/- 0.04, N = 3 SE +/- 0.13, N = 3 SE +/- 0.11, N = 3 SE +/- 0.17, N = 3 SE +/- 0.35, N = 3 SE +/- 0.13, N = 3 SE +/- 0.22, N = 3 SE +/- 0.25, N = 3 SE +/- 0.45, N = 3 SE +/- 0.89, N = 3 SE +/- 0.84, N = 3 SE +/- 1.74, N = 15 SE +/- 3.87, N = 3 16.55 21.34 22.59 23.66 24.55 25.98 30.47 31.79 33.50 34.16 34.25 34.31 36.61 38.10 40.63 66.44 66.63 83.77 146.78 160.13 266.10 285.99 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OpenVINO This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Handwritten English Recognition FP16-INT8 - Device: CPU Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Max 9480 Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9684X EPYC 9684X 2P EPYC 9654 Xeon Platinum 8490H EPYC 9654 2P EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 40 80 120 160 200 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 SE +/- 0.05, N = 3 SE +/- 0.00, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 SE +/- 0.07, N = 3 SE +/- 0.06, N = 3 SE +/- 0.04, N = 3 SE +/- 0.07, N = 3 SE +/- 0.09, N = 3 SE +/- 0.08, N = 3 SE +/- 0.09, N = 3 SE +/- 0.02, N = 3 SE +/- 0.45, N = 15 SE +/- 0.46, N = 3 SE +/- 0.10, N = 3 SE +/- 0.33, N = 3 SE +/- 0.34, N = 3 SE +/- 0.13, N = 3 21.87 22.96 23.12 24.00 24.94 25.12 25.72 32.00 33.18 34.92 36.20 36.90 38.85 39.86 41.38 45.71 60.90 62.64 153.76 157.28 172.72 180.29 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2024.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E Xeon Platinum 8592+ EPYC 9754 Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9654 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.48 / Avg: 135.3 / Max: 146.77 Min: 17.94 / Avg: 165.08 / Max: 177.71 Min: 43.65 / Avg: 235.17 / Max: 270.9 Min: 49.49 / Avg: 255.74 / Max: 292.14 Min: 43.62 / Avg: 308.92 / Max: 356.46 Min: 24.83 / Avg: 318.97 / Max: 381.52 Min: 23.52 / Avg: 323.53 / Max: 352.29 Min: 59.81 / Avg: 328.72 / Max: 377.28 Min: 75.49 / Avg: 331.88 / Max: 373.56 Min: 87.21 / Avg: 332.74 / Max: 375.71 Min: 30.45 / Avg: 333.47 / Max: 360.29 Min: 31.71 / Avg: 382.65 / Max: 411.13 Min: 106.76 / Avg: 462.61 / Max: 540.57 Min: 102.66 / Avg: 511.4 / Max: 581.89 Min: 108.82 / Avg: 608.93 / Max: 712.7 Min: 42.75 / Avg: 632.45 / Max: 690.87 Min: 59.67 / Avg: 652.58 / Max: 705.65 Min: 79.4 / Avg: 653.57 / Max: 770.7 Min: 175.49 / Avg: 660.59 / Max: 741.22 Min: 200.26 / Avg: 664.69 / Max: 755.11 Min: 208.68 / Avg: 665.43 / Max: 751.81 Min: 67.9 / Avg: 760.8 / Max: 819.21
OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Handwritten English Recognition FP16-INT8 - Device: CPU EPYC 9754 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9754 EPYC 9684X Xeon Platinum 8592+ EPYC 9654 Xeon Max 9480 Xeon Max 9468 Xeon 6780E 2P Xeon Platinum 8490H Xeon 6766E 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon 6780E Xeon 6766E Xeon Platinum 8380 1400 2800 4200 5600 7000 SE +/- 14.68, N = 3 SE +/- 13.59, N = 3 SE +/- 3.43, N = 3 SE +/- 9.83, N = 3 SE +/- 2.46, N = 3 SE +/- 2.81, N = 3 SE +/- 1.09, N = 3 SE +/- 7.27, N = 3 SE +/- 4.06, N = 3 SE +/- 0.73, N = 3 SE +/- 4.84, N = 3 SE +/- 4.55, N = 3 SE +/- 3.41, N = 3 SE +/- 3.81, N = 3 SE +/- 1.46, N = 3 SE +/- 1.20, N = 3 SE +/- 3.51, N = 3 SE +/- 0.47, N = 3 SE +/- 9.58, N = 15 SE +/- 0.61, N = 3 SE +/- 1.73, N = 3 SE +/- 4.60, N = 3 6400.94 5766.62 5570.63 5178.29 4744.05 4351.59 3998.02 3280.25 2986.33 2924.66 2739.01 2243.84 2074.30 1826.82 1643.43 1593.88 1542.61 1396.78 1313.39 934.97 832.29 637.76 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
Result
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Vehicle Detection FP16-INT8 - Device: CPU EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9684X EPYC 9654 Xeon Platinum 8380 EPYC 8534P EPYC 9754 EPYC 8534PN Xeon Platinum 8592+ Xeon 6780E Xeon 6780E 2P Xeon Platinum 8592+ 2P Xeon 6766E Xeon 6766E 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9480 2P Xeon Platinum 8490H 2P 5 10 15 20 25 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.12, N = 15 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 4.08 4.46 4.54 5.00 7.85 8.58 9.26 10.78 10.97 11.48 13.73 14.01 14.30 14.67 15.11 16.28 16.33 17.19 17.84 19.44 20.24 21.73 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2024.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E Xeon Platinum 8592+ EPYC 9754 EPYC 9654 Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.95 / Avg: 132.64 / Max: 143.89 Min: 17.62 / Avg: 156.89 / Max: 169.02 Min: 43.79 / Avg: 234.55 / Max: 257.28 Min: 49.67 / Avg: 255.11 / Max: 289.85 Min: 43.53 / Avg: 284.95 / Max: 316.24 Min: 33.71 / Avg: 302.37 / Max: 379.73 Min: 21.25 / Avg: 305.35 / Max: 334.91 Min: 29.58 / Avg: 315.71 / Max: 340.19 Min: 59.46 / Avg: 330.49 / Max: 383.24 Min: 71.96 / Avg: 331 / Max: 377.8 Min: 75.87 / Avg: 332.1 / Max: 380.66 Min: 34.99 / Avg: 371.14 / Max: 399.65 Min: 107.07 / Avg: 468.4 / Max: 537.67 Min: 100.89 / Avg: 510.45 / Max: 581 Min: 110.43 / Avg: 596.74 / Max: 654.1 Min: 43.09 / Avg: 614.49 / Max: 665.93 Min: 60.42 / Avg: 615.26 / Max: 666.72 Min: 85.37 / Avg: 650.24 / Max: 758.93 Min: 169.19 / Avg: 657.97 / Max: 750.83 Min: 209.95 / Avg: 663.67 / Max: 754.1 Min: 203.39 / Avg: 666.38 / Max: 758.84 Min: 68.28 / Avg: 749.07 / Max: 815.77
OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Vehicle Detection FP16-INT8 - Device: CPU EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9754 Xeon Max 9468 2P EPYC 9654 Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon 6766E 2P Xeon Platinum 8490H EPYC 8534P Xeon Max 9468 Xeon Max 9480 EPYC 8534PN Xeon 6780E Xeon 6766E Xeon Platinum 8380 3K 6K 9K 12K 15K SE +/- 1.24, N = 3 SE +/- 4.04, N = 3 SE +/- 4.21, N = 3 SE +/- 6.25, N = 3 SE +/- 3.54, N = 3 SE +/- 66.33, N = 15 SE +/- 4.11, N = 3 SE +/- 8.39, N = 3 SE +/- 8.70, N = 3 SE +/- 11.51, N = 3 SE +/- 5.08, N = 3 SE +/- 0.73, N = 3 SE +/- 2.84, N = 3 SE +/- 8.74, N = 3 SE +/- 0.81, N = 3 SE +/- 2.22, N = 3 SE +/- 8.09, N = 3 SE +/- 5.68, N = 3 SE +/- 2.89, N = 3 SE +/- 2.61, N = 3 SE +/- 1.12, N = 3 SE +/- 1.06, N = 3 12750.59 11735.42 10708.91 8719.37 6017.19 5778.51 5576.86 5528.15 5525.56 5483.81 4992.78 4656.92 4390.81 4376.09 3325.90 2943.08 2935.47 2877.32 2763.48 2553.55 2366.83 2120.33 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9654 EPYC 9684X Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9754 2P Xeon 6780E Xeon Platinum 8380 2P EPYC 9654 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 13 26 39 52 65 SE +/- 0.01, N = 3 SE +/- 0.07, N = 3 SE +/- 0.00, N = 3 SE +/- 0.06, N = 3 SE +/- 0.41, N = 3 SE +/- 0.22, N = 3 SE +/- 0.13, N = 3 SE +/- 0.08, N = 3 SE +/- 0.30, N = 3 SE +/- 0.18, N = 3 SE +/- 0.33, N = 3 SE +/- 0.43, N = 4 SE +/- 0.35, N = 7 SE +/- 0.07, N = 3 SE +/- 0.07, N = 3 SE +/- 0.07, N = 3 SE +/- 0.55, N = 3 SE +/- 0.32, N = 3 SE +/- 0.15, N = 3 SE +/- 0.05, N = 3 SE +/- 0.12, N = 3 SE +/- 0.71, N = 3 29.96 30.92 31.53 33.29 33.52 34.81 35.07 35.11 35.60 35.79 35.88 36.45 38.00 38.42 38.52 39.02 40.45 40.52 41.57 42.90 54.25 56.77 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P 80 160 240 320 400 Min: 16.17 / Avg: 95.68 / Max: 104.17 Min: 17.92 / Avg: 105.97 / Max: 115.61 Min: 43.71 / Avg: 132.85 / Max: 145.13 Min: 43.46 / Avg: 155.26 / Max: 169.97 Min: 21.43 / Avg: 160.48 / Max: 176.89 Min: 29.71 / Avg: 198.61 / Max: 217.44 Min: 53.88 / Avg: 210.59 / Max: 230.53 Min: 33.52 / Avg: 227.41 / Max: 249.95 Min: 104.02 / Avg: 236.43 / Max: 258.12 Min: 106.52 / Avg: 256.16 / Max: 280.65 Min: 41.16 / Avg: 302.54 / Max: 329.11 Min: 101.44 / Avg: 386.17 / Max: 431.52 Min: 58.29 / Avg: 391.03 / Max: 427.27 Min: 65.93 / Avg: 426.21 / Max: 468.57
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9654 EPYC 9684X Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9754 2P Xeon 6780E Xeon Platinum 8380 2P EPYC 9654 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 8 16 24 32 40 SE +/- 0.01, N = 3 SE +/- 0.07, N = 3 SE +/- 0.00, N = 3 SE +/- 0.06, N = 3 SE +/- 0.36, N = 3 SE +/- 0.19, N = 3 SE +/- 0.10, N = 3 SE +/- 0.07, N = 3 SE +/- 0.24, N = 3 SE +/- 0.14, N = 3 SE +/- 0.25, N = 3 SE +/- 0.33, N = 4 SE +/- 0.24, N = 7 SE +/- 0.05, N = 3 SE +/- 0.05, N = 3 SE +/- 0.05, N = 3 SE +/- 0.34, N = 3 SE +/- 0.19, N = 3 SE +/- 0.08, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.22, N = 3 33.38 32.34 31.71 30.04 29.84 28.73 28.52 28.48 28.09 27.94 27.87 27.45 26.33 26.03 25.96 25.63 24.73 24.68 24.05 23.31 18.43 17.62 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9654 EPYC 9684X Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9754 2P Xeon 6780E Xeon Platinum 8380 2P EPYC 9654 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 13 26 39 52 65 SE +/- 0.01, N = 3 SE +/- 0.07, N = 3 SE +/- 0.00, N = 3 SE +/- 0.06, N = 3 SE +/- 0.41, N = 3 SE +/- 0.22, N = 3 SE +/- 0.13, N = 3 SE +/- 0.08, N = 3 SE +/- 0.30, N = 3 SE +/- 0.18, N = 3 SE +/- 0.33, N = 3 SE +/- 0.43, N = 4 SE +/- 0.35, N = 7 SE +/- 0.07, N = 3 SE +/- 0.07, N = 3 SE +/- 0.07, N = 3 SE +/- 0.55, N = 3 SE +/- 0.32, N = 3 SE +/- 0.15, N = 3 SE +/- 0.05, N = 3 SE +/- 0.12, N = 3 SE +/- 0.71, N = 3 29.96 30.92 31.53 33.29 33.52 34.81 35.07 35.11 35.60 35.79 35.88 36.45 38.00 38.42 38.52 39.02 40.45 40.52 41.57 42.90 54.25 56.77 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 512 - Model: AlexNet Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P Xeon Max 9480 2P EPYC 9754 Xeon Max 9468 2P Xeon Platinum 8592+ Xeon Platinum 8380 2P EPYC 9684X EPYC 9654 Xeon Platinum 8490H Xeon Max 9480 Xeon 6780E 2P Xeon Max 9468 Xeon 6766E 2P Xeon Platinum 8380 EPYC 8534P EPYC 8534PN Xeon 6780E Xeon 6766E 400 800 1200 1600 2000 SE +/- 19.23, N = 3 SE +/- 18.52, N = 3 SE +/- 23.15, N = 4 SE +/- 7.97, N = 3 SE +/- 12.71, N = 3 SE +/- 8.55, N = 3 SE +/- 4.07, N = 3 SE +/- 17.88, N = 4 SE +/- 1.84, N = 3 SE +/- 5.91, N = 3 SE +/- 0.49, N = 3 SE +/- 0.67, N = 3 SE +/- 0.51, N = 3 SE +/- 5.13, N = 3 SE +/- 38.99, N = 12 SE +/- 3.89, N = 3 SE +/- 20.10, N = 12 SE +/- 1.47, N = 3 SE +/- 0.35, N = 3 SE +/- 0.57, N = 3 SE +/- 0.86, N = 3 SE +/- 2.03, N = 3 2039.63 1899.64 1897.64 1876.66 1862.79 1677.62 1620.53 1604.11 1486.12 1470.46 1392.91 1363.33 1278.45 1108.95 1064.18 1047.96 979.89 883.00 875.23 796.82 732.79 677.40 images/sec Per Watt
OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 512 - Model: AlexNet EPYC 8534PN EPYC 9754 EPYC 8534P EPYC 9654 Xeon Platinum 8592+ EPYC 9684X EPYC 9754 2P Xeon Platinum 8490H Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 Xeon 6766E Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon 6780E Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6766E 2P Xeon 6780E 2P 2 4 6 8 10 6.990 6.899 6.494 5.294 4.861 4.675 4.569 4.426 3.903 3.853 3.526 3.517 3.502 3.409 3.220 3.147 3.101 3.050 2.674 2.537 2.406 2.310 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.16.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E Xeon Platinum 8380 EPYC 9654 Xeon Platinum 8490H EPYC 9684X Xeon Platinum 8592+ Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 120 240 360 480 600 Min: 17.15 / Avg: 114 / Max: 136.39 Min: 19.53 / Avg: 134.77 / Max: 159.36 Min: 43.65 / Avg: 193.42 / Max: 225.02 Min: 21.12 / Avg: 234.91 / Max: 276.8 Min: 43.52 / Avg: 240.23 / Max: 284.38 Min: 53.82 / Avg: 251.06 / Max: 280.69 Min: 29.56 / Avg: 257.54 / Max: 304.86 Min: 60.6 / Avg: 288.85 / Max: 359.42 Min: 32.89 / Avg: 297.93 / Max: 351.95 Min: 33.58 / Avg: 305.73 / Max: 353.3 Min: 83.71 / Avg: 325.32 / Max: 365.52 Min: 73.05 / Avg: 325.43 / Max: 363.64 Min: 105.98 / Avg: 407.31 / Max: 478.77 Min: 42.21 / Avg: 410.73 / Max: 503.82 Min: 108.97 / Avg: 460.62 / Max: 565.85 Min: 99.17 / Avg: 474.18 / Max: 540.4 Min: 59.19 / Avg: 483.44 / Max: 576.3 Min: 84.88 / Avg: 522.57 / Max: 631.68 Min: 64.96 / Avg: 538.13 / Max: 644.98 Min: 161.55 / Avg: 603.65 / Max: 700.57 Min: 209.29 / Avg: 625.4 / Max: 706.28 Min: 200.65 / Avg: 632.22 / Max: 705.95
QuantLib QuantLib is an open-source library/framework around quantitative finance for modeling, trading and risk management scenarios. QuantLib is written in C++ with Boost and its built-in benchmark used reports the QuantLib Benchmark Index benchmark score. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org MFLOPS, More Is Better QuantLib 1.32 Configuration: Multi-Threaded EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8592+ 2P EPYC 9754 Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9480 2P EPYC 9654 Xeon Max 9468 2P Xeon 6780E Xeon 6766E Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H EPYC 8534P Xeon Max 9480 EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 160K 320K 480K 640K 800K SE +/- 804.57, N = 3 SE +/- 1065.76, N = 3 SE +/- 713.18, N = 3 SE +/- 1647.70, N = 3 SE +/- 1610.78, N = 3 SE +/- 248.00, N = 3 SE +/- 156.35, N = 3 SE +/- 362.38, N = 3 SE +/- 412.72, N = 3 SE +/- 160.59, N = 3 SE +/- 86.78, N = 3 SE +/- 286.86, N = 3 SE +/- 846.02, N = 3 SE +/- 306.14, N = 3 SE +/- 63.14, N = 3 SE +/- 457.00, N = 3 SE +/- 105.60, N = 3 SE +/- 82.80, N = 3 SE +/- 40.48, N = 3 SE +/- 86.28, N = 3 SE +/- 155.19, N = 3 SE +/- 395.04, N = 3 759922.1 644091.0 593246.9 519392.2 456271.0 380867.5 374413.5 350187.5 319827.2 316359.2 294183.2 288129.1 262432.6 233836.3 195648.8 182366.9 179647.5 172566.8 162727.4 153920.8 143755.3 91036.5 1. (CXX) g++ options: -O3 -march=native -fPIE -pie MFLOPS Per Watt
OpenBenchmarking.org MFLOPS Per Watt, More Is Better QuantLib 1.32 Configuration: Multi-Threaded EPYC 8534PN EPYC 9754 EPYC 9754 2P Xeon 6766E EPYC 8534P Xeon 6766E 2P Xeon 6780E Xeon 6780E 2P EPYC 9654 EPYC 9654 2P EPYC 9684X 2P EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Platinum 8380 300 600 900 1200 1500 1202.80 1200.71 1200.40 1113.28 1092.36 1089.21 960.37 956.27 930.70 921.01 881.58 864.58 605.78 585.75 552.74 537.18 495.61 477.83 447.64 436.19 359.31 358.69 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better QuantLib 1.32 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 EPYC 9654 Xeon Max 9468 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.19 / Avg: 127.97 / Max: 140.32 Min: 17.59 / Avg: 157.98 / Max: 175.08 Min: 43.59 / Avg: 210.04 / Max: 270.15 Min: 49.64 / Avg: 253.81 / Max: 290.36 Min: 43.42 / Avg: 273.26 / Max: 357.57 Min: 21.13 / Avg: 311.83 / Max: 350.73 Min: 28.93 / Avg: 316.09 / Max: 352.6 Min: 72.72 / Avg: 321.14 / Max: 375.61 Min: 31.57 / Avg: 322.97 / Max: 378.63 Min: 58.59 / Avg: 325.01 / Max: 377.73 Min: 83.08 / Avg: 328.34 / Max: 376.88 Min: 33.64 / Avg: 369.92 / Max: 411.04 Min: 103.76 / Avg: 418.9 / Max: 541.48 Min: 100.73 / Avg: 507.54 / Max: 581.6 Min: 104.62 / Avg: 543.14 / Max: 713.12 Min: 40.73 / Avg: 633.06 / Max: 715.84 Min: 55.66 / Avg: 644.13 / Max: 715.22 Min: 77.95 / Avg: 650.23 / Max: 756.48 Min: 167.37 / Avg: 651.9 / Max: 756.65 Min: 205.33 / Avg: 660.57 / Max: 754.82 Min: 203.63 / Avg: 662.07 / Max: 755.13 Min: 65.34 / Avg: 730.61 / Max: 822.15
OpenVINO This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Face Detection FP16-INT8 - Device: CPU Xeon Platinum 8380 2P Xeon Platinum 8592+ EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Max 9468 EPYC 9654 2P Xeon Max 9468 2P Xeon Max 9480 EPYC 9754 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Platinum 8380 EPYC 9684X EPYC 9654 EPYC 9754 EPYC 8534P Xeon 6780E Xeon 6780E 2P EPYC 8534PN Xeon 6766E Xeon 6766E 2P 160 320 480 640 800 SE +/- 0.22, N = 3 SE +/- 0.69, N = 3 SE +/- 0.43, N = 3 SE +/- 0.33, N = 3 SE +/- 0.15, N = 3 SE +/- 0.02, N = 3 SE +/- 0.40, N = 3 SE +/- 0.47, N = 3 SE +/- 0.03, N = 3 SE +/- 0.35, N = 3 SE +/- 0.00, N = 3 SE +/- 0.28, N = 3 SE +/- 3.13, N = 3 SE +/- 0.43, N = 3 SE +/- 0.07, N = 3 SE +/- 0.18, N = 3 SE +/- 0.18, N = 3 SE +/- 1.47, N = 3 SE +/- 1.63, N = 3 SE +/- 0.12, N = 3 SE +/- 0.99, N = 3 SE +/- 1.58, N = 3 207.84 224.64 233.63 234.24 236.52 253.63 256.63 257.50 275.53 278.52 283.11 356.38 415.89 471.98 510.10 551.07 592.22 618.97 630.23 654.83 692.79 719.83 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2024.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.81 / Avg: 126.81 / Max: 139.78 Min: 18.22 / Avg: 153 / Max: 168.49 Min: 43.56 / Avg: 225.98 / Max: 269.62 Min: 50.07 / Avg: 251.32 / Max: 290.45 Min: 20.35 / Avg: 295.9 / Max: 332.78 Min: 43.45 / Avg: 297.23 / Max: 353.99 Min: 28.86 / Avg: 317.58 / Max: 349.47 Min: 35.2 / Avg: 319.39 / Max: 384.58 Min: 58.75 / Avg: 323.68 / Max: 377.26 Min: 73.05 / Avg: 326.77 / Max: 376.97 Min: 82.85 / Avg: 327.08 / Max: 376.47 Min: 35.31 / Avg: 374.05 / Max: 411.15 Min: 105.04 / Avg: 434 / Max: 540.8 Min: 102.18 / Avg: 501.41 / Max: 578.17 Min: 107.91 / Avg: 568.65 / Max: 711.11 Min: 42.88 / Avg: 582.16 / Max: 663.93 Min: 74.99 / Avg: 625.11 / Max: 757.77 Min: 53.33 / Avg: 629.45 / Max: 695.56 Min: 166.52 / Avg: 639.29 / Max: 750.58 Min: 200.71 / Avg: 649.1 / Max: 748.61 Min: 204.33 / Avg: 649.97 / Max: 743.7 Min: 67.22 / Avg: 740.61 / Max: 819.79
OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Face Detection FP16-INT8 - Device: CPU Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 9754 2P Xeon Max 9480 EPYC 9684X 2P Xeon Max 9468 EPYC 9654 2P Xeon Platinum 8490H EPYC 9754 Xeon 6780E 2P EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P EPYC 9654 Xeon 6780E EPYC 8534P Xeon 6766E EPYC 8534PN Xeon Platinum 8380 120 240 360 480 600 SE +/- 0.77, N = 3 SE +/- 0.01, N = 3 SE +/- 0.54, N = 3 SE +/- 0.58, N = 3 SE +/- 0.69, N = 3 SE +/- 0.04, N = 3 SE +/- 0.41, N = 3 SE +/- 0.41, N = 3 SE +/- 0.14, N = 3 SE +/- 0.01, N = 3 SE +/- 0.12, N = 3 SE +/- 0.03, N = 3 SE +/- 0.18, N = 3 SE +/- 0.12, N = 3 SE +/- 0.18, N = 3 SE +/- 0.10, N = 3 SE +/- 0.01, N = 3 SE +/- 0.15, N = 3 SE +/- 0.01, N = 3 SE +/- 0.07, N = 3 SE +/- 0.06, N = 3 SE +/- 0.36, N = 3 545.21 423.18 401.45 373.50 284.64 231.83 217.11 205.03 202.59 188.74 167.83 115.78 113.68 101.51 99.72 96.08 93.99 58.06 53.89 51.85 48.73 47.99 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OSPRay Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Items Per Second, More Is Better OSPRay 3.1 Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_time EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9754 Xeon Max 9468 2P EPYC 9684X EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6780E 2P Xeon Platinum 8380 2P Xeon Max 9480 Xeon 6766E 2P Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon 6780E Xeon 6766E Xeon Platinum 8380 14 28 42 56 70 SE +/- 0.05, N = 3 SE +/- 0.14, N = 3 SE +/- 0.07, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 SE +/- 0.16, N = 3 SE +/- 0.02, N = 3 SE +/- 0.09, N = 3 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.23, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.04, N = 3 63.85 60.37 52.67 48.76 42.82 39.34 37.74 35.93 34.64 30.61 27.10 24.64 23.12 22.51 22.36 21.22 19.91 18.47 16.92 13.40 12.17 11.86 Items Per Second Per Watt
OpenBenchmarking.org Items Per Second Per Watt, More Is Better OSPRay 3.1 Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_time EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 EPYC 9754 2P EPYC 9684X EPYC 9654 2P Xeon Platinum 8592+ EPYC 9684X 2P Xeon 6766E Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon 6780E Xeon Platinum 8490H 2P Xeon Max 9468 Xeon 6766E 2P Xeon Max 9480 2P Xeon 6780E 2P Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 0.032 0.064 0.096 0.128 0.16 0.142 0.135 0.130 0.112 0.111 0.094 0.091 0.088 0.084 0.079 0.079 0.077 0.070 0.069 0.068 0.063 0.061 0.060 0.058 0.055 0.049 0.047 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OSPRay 3.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9654 EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.8 / Avg: 119.14 / Max: 138.68 Min: 18.46 / Avg: 136.83 / Max: 158.83 Min: 43.64 / Avg: 154.91 / Max: 191.69 Min: 43.52 / Avg: 195.57 / Max: 243.19 Min: 53.09 / Avg: 241.12 / Max: 287.29 Min: 29.59 / Avg: 273.92 / Max: 319.11 Min: 23.6 / Avg: 289.58 / Max: 332.39 Min: 26.13 / Avg: 308.08 / Max: 376.91 Min: 61.9 / Avg: 311.6 / Max: 375.6 Min: 76.02 / Avg: 317.76 / Max: 373.98 Min: 78.8 / Avg: 318.09 / Max: 375.14 Min: 105.18 / Avg: 346.87 / Max: 441 Min: 35.77 / Avg: 367.12 / Max: 410.2 Min: 107.82 / Avg: 395.24 / Max: 502.47 Min: 100.97 / Avg: 476.45 / Max: 565.54 Min: 42.2 / Avg: 573.28 / Max: 670.75 Min: 60.47 / Avg: 577.89 / Max: 656.09 Min: 77.1 / Avg: 630.9 / Max: 750.84 Min: 174.41 / Avg: 631.92 / Max: 708.59 Min: 207.37 / Avg: 650.92 / Max: 752.3 Min: 206.3 / Avg: 652.41 / Max: 753.05 Min: 65.72 / Avg: 718.49 / Max: 821.14
Blender Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.1 Blend File: Pabellon Barcelona - Compute: CPU-Only EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 Xeon Platinum 8592+ 2P EPYC 9684X Xeon Platinum 8490H 2P EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E Xeon Platinum 8380 2P Xeon Platinum 8592+ EPYC 8534P Xeon Platinum 8490H EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 30 60 90 120 150 SE +/- 0.13, N = 3 SE +/- 0.11, N = 3 SE +/- 0.10, N = 3 SE +/- 0.17, N = 3 SE +/- 0.29, N = 3 SE +/- 0.13, N = 3 SE +/- 0.06, N = 3 SE +/- 0.16, N = 3 SE +/- 0.26, N = 3 SE +/- 0.04, N = 3 SE +/- 0.20, N = 3 SE +/- 0.08, N = 3 SE +/- 0.25, N = 3 SE +/- 0.13, N = 3 SE +/- 0.26, N = 3 SE +/- 0.09, N = 3 SE +/- 0.16, N = 3 SE +/- 0.15, N = 3 SE +/- 0.05, N = 3 SE +/- 0.44, N = 3 SE +/- 0.31, N = 3 SE +/- 0.15, N = 3 21.66 23.84 26.35 31.28 35.11 39.83 41.85 44.87 46.38 48.88 51.29 55.26 60.11 65.67 76.10 80.29 85.12 87.92 94.41 97.80 106.96 150.39 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Blender 4.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9684X Xeon 6766E 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.1 / Avg: 133.17 / Max: 140.81 Min: 19.07 / Avg: 163.93 / Max: 175 Min: 43.71 / Avg: 217.74 / Max: 239.54 Min: 52.51 / Avg: 261.74 / Max: 291.09 Min: 43.58 / Avg: 280.84 / Max: 315.47 Min: 21.64 / Avg: 304.63 / Max: 348.18 Min: 30.54 / Avg: 312.47 / Max: 346.36 Min: 34.2 / Avg: 330.4 / Max: 377.77 Min: 60.61 / Avg: 332.25 / Max: 377.75 Min: 76.03 / Avg: 336.64 / Max: 376.9 Min: 76.38 / Avg: 337.24 / Max: 375.65 Min: 31.38 / Avg: 370.95 / Max: 410.68 Min: 105.22 / Avg: 415.82 / Max: 498.21 Min: 41.28 / Avg: 465.54 / Max: 700.45 Min: 98.03 / Avg: 509.42 / Max: 580.91 Min: 108.3 / Avg: 523.16 / Max: 632.46 Min: 58.1 / Avg: 592.5 / Max: 710.39 Min: 77.37 / Avg: 624.76 / Max: 738.53 Min: 172.77 / Avg: 638.84 / Max: 757.46 Min: 208.34 / Avg: 654.26 / Max: 753.55 Min: 211.91 / Avg: 656.12 / Max: 755.31 Min: 65.23 / Avg: 677.05 / Max: 822.1
OpenVINO This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Person Vehicle Bike Detection FP16 - Device: CPU EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P EPYC 9684X EPYC 9654 Xeon Platinum 8380 2P Xeon Platinum 8592+ EPYC 8534P EPYC 9754 Xeon Platinum 8592+ 2P Xeon Max 9468 EPYC 8534PN Xeon Max 9468 2P Xeon Max 9480 Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Platinum 8380 Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 6 12 18 24 30 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.11, N = 3 SE +/- 0.00, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.00, N = 3 SE +/- 0.03, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.00, N = 3 SE +/- 0.05, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.05, N = 3 4.91 5.20 6.27 8.48 9.47 9.82 11.17 11.22 11.35 11.69 12.54 12.67 13.34 13.59 14.42 15.04 16.50 18.08 23.49 24.24 25.33 26.12 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2024.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9480 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 15.14 / Avg: 128.27 / Max: 140.82 Min: 18.03 / Avg: 158.07 / Max: 171.2 Min: 43.71 / Avg: 214.7 / Max: 232.23 Min: 50.78 / Avg: 254.29 / Max: 289.74 Min: 43.7 / Avg: 257.32 / Max: 280.77 Min: 22.15 / Avg: 293.51 / Max: 329.03 Min: 29.5 / Avg: 315.59 / Max: 344.89 Min: 33.49 / Avg: 325.49 / Max: 379.56 Min: 73.58 / Avg: 327.03 / Max: 377.73 Min: 60.78 / Avg: 328.2 / Max: 380.56 Min: 83.78 / Avg: 329.87 / Max: 374.74 Min: 34.6 / Avg: 368.08 / Max: 403.09 Min: 107.91 / Avg: 444.05 / Max: 496.43 Min: 101.65 / Avg: 506.94 / Max: 559.01 Min: 110.24 / Avg: 513.09 / Max: 584.23 Min: 43.98 / Avg: 604.36 / Max: 672.48 Min: 35.7 / Avg: 634.64 / Max: 688.41 Min: 84.22 / Avg: 639.31 / Max: 759.19 Min: 174.42 / Avg: 651.12 / Max: 753.13 Min: 204.78 / Avg: 655.73 / Max: 753.37 Min: 206.04 / Avg: 658.8 / Max: 752.52 Min: 70.55 / Avg: 710.38 / Max: 794.96
OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Person Vehicle Bike Detection FP16 - Device: CPU Xeon Platinum 8592+ 2P EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 9684X EPYC 9754 EPYC 9654 Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H Xeon 6780E 2P EPYC 8534P Xeon 6766E 2P EPYC 8534PN Xeon Platinum 8380 2P Xeon 6780E Xeon 6766E Xeon Platinum 8380 2K 4K 6K 8K 10K SE +/- 3.37, N = 3 SE +/- 4.91, N = 3 SE +/- 10.88, N = 3 SE +/- 6.56, N = 3 SE +/- 8.98, N = 3 SE +/- 3.66, N = 3 SE +/- 4.05, N = 3 SE +/- 4.97, N = 3 SE +/- 18.02, N = 3 SE +/- 54.09, N = 3 SE +/- 25.68, N = 3 SE +/- 9.37, N = 3 SE +/- 7.17, N = 3 SE +/- 4.52, N = 3 SE +/- 3.86, N = 3 SE +/- 2.97, N = 3 SE +/- 5.02, N = 3 SE +/- 3.87, N = 3 SE +/- 1.00, N = 3 SE +/- 2.87, N = 3 SE +/- 2.55, N = 3 SE +/- 0.24, N = 3 10934.17 10182.16 9752.89 9203.19 7862.58 7689.32 7127.28 5725.03 5608.07 5594.81 5024.32 4093.73 3802.61 3606.12 2957.69 2830.12 2743.39 2506.10 2030.97 1525.26 1414.86 1100.18 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: bertsquad-12 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X EPYC 9654 EPYC 8534P EPYC 9754 Xeon 6780E EPYC 8534PN EPYC 9654 2P EPYC 9684X 2P Xeon 6766E EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 30 60 90 120 150 SE +/- 0.57, N = 3 SE +/- 0.49, N = 3 SE +/- 0.23, N = 3 SE +/- 0.18, N = 3 SE +/- 0.31, N = 3 SE +/- 0.12, N = 3 SE +/- 0.56, N = 3 SE +/- 0.11, N = 3 SE +/- 0.50, N = 3 SE +/- 0.23, N = 3 SE +/- 0.57, N = 3 SE +/- 0.60, N = 3 SE +/- 1.08, N = 3 SE +/- 0.60, N = 3 SE +/- 0.31, N = 3 SE +/- 1.29, N = 4 SE +/- 0.65, N = 3 SE +/- 0.60, N = 3 SE +/- 0.10, N = 3 SE +/- 0.30, N = 3 SE +/- 0.21, N = 3 SE +/- 0.90, N = 3 50.83 59.94 61.25 62.13 63.54 64.79 66.91 68.83 72.80 73.97 81.43 81.60 102.05 103.60 105.22 112.76 113.52 114.16 116.64 125.24 140.07 149.99 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P 90 180 270 360 450 Min: 16.6 / Avg: 99.26 / Max: 116.55 Min: 18.28 / Avg: 121.13 / Max: 131.61 Min: 43.69 / Avg: 144.23 / Max: 154.04 Min: 43.61 / Avg: 168.62 / Max: 180.75 Min: 20.48 / Avg: 190.15 / Max: 204.22 Min: 30.08 / Avg: 223.68 / Max: 241.02 Min: 53.86 / Avg: 240.25 / Max: 259.56 Min: 33.93 / Avg: 264.24 / Max: 286.38 Min: 104.24 / Avg: 292.81 / Max: 313.39 Min: 40.61 / Avg: 307.67 / Max: 327.17 Min: 105.38 / Avg: 317.3 / Max: 341.18 Min: 58.35 / Avg: 411.86 / Max: 441.41 Min: 103.61 / Avg: 436.17 / Max: 466.43 Min: 68.13 / Avg: 462.27 / Max: 495
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: bertsquad-12 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X EPYC 9654 EPYC 8534P EPYC 9754 Xeon 6780E EPYC 8534PN EPYC 9654 2P EPYC 9684X 2P Xeon 6766E EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 5 10 15 20 25 SE +/- 0.21736, N = 3 SE +/- 0.13407, N = 3 SE +/- 0.06023, N = 3 SE +/- 0.04566, N = 3 SE +/- 0.07567, N = 3 SE +/- 0.02832, N = 3 SE +/- 0.12404, N = 3 SE +/- 0.02421, N = 3 SE +/- 0.09371, N = 3 SE +/- 0.04124, N = 3 SE +/- 0.08493, N = 3 SE +/- 0.09004, N = 3 SE +/- 0.10485, N = 3 SE +/- 0.05587, N = 3 SE +/- 0.02804, N = 3 SE +/- 0.10146, N = 4 SE +/- 0.05052, N = 3 SE +/- 0.04604, N = 3 SE +/- 0.00728, N = 3 SE +/- 0.01907, N = 3 SE +/- 0.01071, N = 3 SE +/- 0.03979, N = 3 19.67730 16.68630 16.32510 16.09560 15.73940 15.43410 14.94600 14.52800 13.73660 13.51900 12.28070 12.25490 9.80090 9.65296 9.50353 8.87144 8.80930 8.75971 8.57286 7.98402 7.13917 6.66718 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: bertsquad-12 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X EPYC 9654 EPYC 8534P EPYC 9754 Xeon 6780E EPYC 8534PN EPYC 9654 2P EPYC 9684X 2P Xeon 6766E EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 30 60 90 120 150 SE +/- 0.57, N = 3 SE +/- 0.49, N = 3 SE +/- 0.23, N = 3 SE +/- 0.18, N = 3 SE +/- 0.31, N = 3 SE +/- 0.12, N = 3 SE +/- 0.56, N = 3 SE +/- 0.11, N = 3 SE +/- 0.50, N = 3 SE +/- 0.23, N = 3 SE +/- 0.57, N = 3 SE +/- 0.60, N = 3 SE +/- 1.08, N = 3 SE +/- 0.60, N = 3 SE +/- 0.31, N = 3 SE +/- 1.29, N = 4 SE +/- 0.65, N = 3 SE +/- 0.60, N = 3 SE +/- 0.10, N = 3 SE +/- 0.30, N = 3 SE +/- 0.21, N = 3 SE +/- 0.90, N = 3 50.83 59.94 61.25 62.13 63.54 64.79 66.91 68.83 72.80 73.97 81.43 81.60 102.05 103.60 105.22 112.76 113.52 114.16 116.64 125.24 140.07 149.99 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OpenVINO This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Person Re-Identification Retail FP16 - Device: CPU EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8592+ 2P Xeon Max 9468 EPYC 9654 Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9480 2P EPYC 9754 EPYC 8534P EPYC 8534PN Xeon Platinum 8490H Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 6 12 18 24 30 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.08, N = 3 SE +/- 0.02, N = 3 SE +/- 0.16, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 3.78 3.98 4.37 6.72 6.93 6.95 7.13 7.23 7.32 7.72 7.76 7.97 8.03 8.59 9.46 9.79 16.36 16.73 17.81 18.75 23.25 24.39 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2024.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9654 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 18.47 / Avg: 133.6 / Max: 149.13 Min: 18.98 / Avg: 159.92 / Max: 172.73 Min: 43.82 / Avg: 230.25 / Max: 249.52 Min: 49.85 / Avg: 255.01 / Max: 291.18 Min: 43.64 / Avg: 297.75 / Max: 328.7 Min: 23.26 / Avg: 322.4 / Max: 353.88 Min: 34.07 / Avg: 323.98 / Max: 378.14 Min: 57.71 / Avg: 328.31 / Max: 374.69 Min: 73.88 / Avg: 330.38 / Max: 377.07 Min: 83.77 / Avg: 330.51 / Max: 377.23 Min: 29.57 / Avg: 331.39 / Max: 358.09 Min: 33.67 / Avg: 380.02 / Max: 410.65 Min: 107.09 / Avg: 456.74 / Max: 515.82 Min: 101.18 / Avg: 504.36 / Max: 573.33 Min: 110.7 / Avg: 588.97 / Max: 656.55 Min: 41.99 / Avg: 631.93 / Max: 700.82 Min: 84.79 / Avg: 638.38 / Max: 770.92 Min: 60.33 / Avg: 641.88 / Max: 698.45 Min: 172.39 / Avg: 651.14 / Max: 755.22 Min: 198.29 / Avg: 659.86 / Max: 754.53 Min: 209.53 / Avg: 660.24 / Max: 754.29 Min: 64.76 / Avg: 750.99 / Max: 819.5
OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Person Re-Identification Retail FP16 - Device: CPU Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9754 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ EPYC 9754 Xeon Max 9480 EPYC 9684X Xeon Max 9468 EPYC 9654 Xeon Platinum 8490H Xeon 6780E 2P Xeon 6766E 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon 6780E Xeon 6766E Xeon Platinum 8380 4K 8K 12K 16K 20K SE +/- 7.74, N = 3 SE +/- 4.14, N = 3 SE +/- 10.13, N = 3 SE +/- 7.22, N = 3 SE +/- 3.67, N = 3 SE +/- 7.99, N = 3 SE +/- 7.03, N = 3 SE +/- 5.57, N = 3 SE +/- 4.98, N = 3 SE +/- 0.46, N = 3 SE +/- 4.58, N = 3 SE +/- 3.72, N = 3 SE +/- 4.14, N = 3 SE +/- 3.97, N = 3 SE +/- 4.90, N = 3 SE +/- 6.55, N = 3 SE +/- 3.18, N = 3 SE +/- 1.89, N = 3 SE +/- 2.06, N = 3 SE +/- 11.75, N = 3 SE +/- 20.57, N = 3 SE +/- 1.47, N = 3 18375.63 15519.92 14601.52 14033.94 13102.28 12665.62 12024.18 9515.86 7862.65 7207.74 6846.79 6717.65 6554.12 5943.66 4259.99 3787.94 3692.67 3355.73 3263.03 2179.99 2006.00 1698.05 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
PyTorch This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org batches/sec, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 1 - Model: ResNet-50 Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P EPYC 9684X EPYC 8534P EPYC 8534PN EPYC 9654 Xeon 6780E Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon 6766E EPYC 9754 EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P 20 40 60 80 100 SE +/- 0.16, N = 3 SE +/- 0.22, N = 3 SE +/- 0.03, N = 3 SE +/- 0.28, N = 3 SE +/- 0.36, N = 3 SE +/- 0.38, N = 8 SE +/- 0.52, N = 3 SE +/- 0.20, N = 3 SE +/- 0.15, N = 3 SE +/- 0.12, N = 3 SE +/- 0.30, N = 3 SE +/- 0.15, N = 3 SE +/- 0.45, N = 4 SE +/- 0.49, N = 3 SE +/- 0.44, N = 3 SE +/- 0.12, N = 3 SE +/- 0.12, N = 3 SE +/- 0.37, N = 12 SE +/- 0.28, N = 15 SE +/- 0.22, N = 4 SE +/- 0.20, N = 3 SE +/- 0.06, N = 3 77.52 61.32 51.46 50.99 48.59 46.49 45.69 42.64 41.17 41.11 40.97 40.45 35.85 35.53 35.46 32.45 31.62 19.84 19.34 17.91 16.45 14.90 batches/sec Per Watt
OpenBenchmarking.org batches/sec Per Watt, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 1 - Model: ResNet-50 EPYC 8534PN EPYC 8534P Xeon Platinum 8592+ Xeon Platinum 8380 Xeon 6766E Xeon 6780E Xeon Platinum 8490H EPYC 9654 EPYC 9754 Xeon Max 9468 EPYC 9684X Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9654 2P Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P EPYC 9684X 2P 0.1026 0.2052 0.3078 0.4104 0.513 0.456 0.435 0.377 0.346 0.258 0.251 0.234 0.231 0.221 0.214 0.200 0.193 0.155 0.102 0.088 0.083 0.066 0.056 0.053 0.053 0.052 0.052 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better PyTorch 2.2.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9754 2P Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 110 220 330 440 550 Min: 16.55 / Avg: 89.88 / Max: 106.44 Min: 17.99 / Avg: 94.53 / Max: 112.2 Min: 43.63 / Avg: 125.95 / Max: 146.47 Min: 43.45 / Avg: 142.9 / Max: 169.81 Min: 21.5 / Avg: 143.05 / Max: 166.54 Min: 30.07 / Avg: 174.93 / Max: 206.71 Min: 52.57 / Avg: 177.33 / Max: 220.98 Min: 33.19 / Avg: 205.45 / Max: 273.87 Min: 33.48 / Avg: 206.21 / Max: 246.32 Min: 58.92 / Avg: 219.54 / Max: 264.14 Min: 73.83 / Avg: 237.82 / Max: 286.68 Min: 78.78 / Avg: 251.12 / Max: 295.23 Min: 101.16 / Avg: 274.76 / Max: 317.28 Min: 41.67 / Avg: 279.95 / Max: 308.17 Min: 105.53 / Avg: 307.69 / Max: 344.16 Min: 108.69 / Avg: 341.29 / Max: 380.81 Min: 58.39 / Avg: 343.94 / Max: 402.47 Min: 62.51 / Avg: 384.75 / Max: 433.08 Min: 160.89 / Avg: 425.75 / Max: 563.56 Min: 81.45 / Avg: 450.14 / Max: 553.23 Min: 204.28 / Avg: 527.85 / Max: 604.75 Min: 199.22 / Avg: 540.93 / Max: 610.22
OpenVINO This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Weld Porosity Detection FP16-INT8 - Device: CPU Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Max 9480 Xeon Max 9480 2P Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon Platinum 8380 EPYC 9684X 2P EPYC 9684X EPYC 9654 EPYC 9654 2P EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 8534PN Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 7 14 21 28 35 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.03, N = 4 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.08, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.04, N = 3 SE +/- 0.05, N = 3 2.36 2.48 2.50 2.50 2.53 2.67 2.79 3.58 8.27 8.58 9.25 9.42 10.09 10.17 10.92 11.20 11.78 13.11 24.25 24.77 27.01 28.24 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2024.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.61 / Avg: 128.88 / Max: 137.57 Min: 18.69 / Avg: 156.99 / Max: 167.88 Min: 43.68 / Avg: 235.37 / Max: 270.07 Min: 50.92 / Avg: 255.35 / Max: 288.21 Min: 21.95 / Avg: 308.04 / Max: 333.19 Min: 43.61 / Avg: 309.55 / Max: 356.77 Min: 28.34 / Avg: 326.4 / Max: 351.05 Min: 34.78 / Avg: 327.29 / Max: 381.26 Min: 58.8 / Avg: 329.61 / Max: 364.22 Min: 73.64 / Avg: 331.47 / Max: 377.08 Min: 83.9 / Avg: 332.66 / Max: 375.98 Min: 34.9 / Avg: 384.48 / Max: 410.25 Min: 106.03 / Avg: 466.01 / Max: 540.12 Min: 100.48 / Avg: 510.62 / Max: 581.76 Min: 108.97 / Avg: 612 / Max: 711.29 Min: 40.68 / Avg: 615.84 / Max: 667.35 Min: 59.07 / Avg: 649.54 / Max: 702.57 Min: 83.6 / Avg: 652.57 / Max: 760.35 Min: 168.91 / Avg: 662.23 / Max: 753.52 Min: 208.01 / Avg: 666.11 / Max: 753.96 Min: 206.25 / Avg: 666.2 / Max: 754.51 Min: 68.11 / Avg: 762.39 / Max: 819.68
OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Weld Porosity Detection FP16-INT8 - Device: CPU Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 9754 2P Xeon Max 9480 EPYC 9684X 2P Xeon Max 9468 EPYC 9654 2P Xeon Platinum 8490H EPYC 9754 Xeon 6780E 2P Xeon 6766E 2P EPYC 9684X Xeon Platinum 8380 2P EPYC 9654 Xeon 6780E EPYC 8534P Xeon 6766E EPYC 8534PN Xeon Platinum 8380 10K 20K 30K 40K 50K SE +/- 39.08, N = 3 SE +/- 53.98, N = 3 SE +/- 233.95, N = 3 SE +/- 80.58, N = 3 SE +/- 47.41, N = 3 SE +/- 11.27, N = 3 SE +/- 21.27, N = 3 SE +/- 110.20, N = 3 SE +/- 234.43, N = 4 SE +/- 147.69, N = 3 SE +/- 2.71, N = 3 SE +/- 5.50, N = 3 SE +/- 19.55, N = 3 SE +/- 12.93, N = 3 SE +/- 11.71, N = 3 SE +/- 6.41, N = 3 SE +/- 4.46, N = 3 SE +/- 6.61, N = 3 SE +/- 0.41, N = 3 SE +/- 7.59, N = 3 SE +/- 1.69, N = 3 SE +/- 5.62, N = 3 48122.16 47654.95 40032.81 37710.60 26943.42 22499.03 20910.53 20550.93 19183.83 18666.16 15943.17 11508.19 11390.04 10020.91 9997.89 9657.08 9356.71 5866.63 5366.79 5278.46 4830.01 4502.74 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: GPT-2 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 EPYC 9684X EPYC 9654 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon 6780E EPYC 9754 Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6766E EPYC 8534PN EPYC 8534P EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 4 8 12 16 20 SE +/- 0.00434, N = 3 SE +/- 0.02114, N = 3 SE +/- 0.01428, N = 3 SE +/- 0.02823, N = 3 SE +/- 0.07483, N = 3 SE +/- 0.01980, N = 3 SE +/- 0.02436, N = 3 SE +/- 0.00351, N = 3 SE +/- 0.06784, N = 3 SE +/- 0.03717, N = 3 SE +/- 0.01181, N = 3 SE +/- 0.00921, N = 3 SE +/- 0.01417, N = 3 SE +/- 0.08283, N = 3 SE +/- 0.00515, N = 3 SE +/- 0.02580, N = 3 SE +/- 0.03809, N = 3 SE +/- 0.03939, N = 3 SE +/- 0.02885, N = 3 SE +/- 0.01997, N = 3 SE +/- 0.01174, N = 3 SE +/- 0.11771, N = 3 4.01687 5.10948 5.21520 5.22781 5.29027 5.54379 6.18478 6.37780 6.50747 6.90619 7.13535 7.28087 7.30822 7.34848 7.59589 7.70511 7.74521 9.06607 9.12452 10.06660 16.54180 17.05530 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8380 EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P 60 120 180 240 300 Min: 15.83 / Avg: 75.69 / Max: 81.06 Min: 16.21 / Avg: 79.24 / Max: 84.5 Min: 43.73 / Avg: 95.8 / Max: 101.7 Min: 43.53 / Avg: 101.74 / Max: 108.09 Min: 19.44 / Avg: 102.38 / Max: 109.98 Min: 27.22 / Avg: 134.62 / Max: 143.75 Min: 35.03 / Avg: 153.85 / Max: 164.75 Min: 103.68 / Avg: 180.17 / Max: 191.33 Min: 94.92 / Avg: 183.21 / Max: 193.59 Min: 53.03 / Avg: 204.41 / Max: 224.2 Min: 40 / Avg: 213.81 / Max: 226.08 Min: 57.53 / Avg: 262.4 / Max: 276.96 Min: 61.69 / Avg: 290.82 / Max: 309 Min: 97.1 / Avg: 306.05 / Max: 325.94
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: GPT-2 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 EPYC 9684X EPYC 9654 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon 6780E EPYC 9754 Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6766E EPYC 8534PN EPYC 8534P EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 50 100 150 200 250 SE +/- 0.27, N = 3 SE +/- 0.81, N = 3 SE +/- 0.53, N = 3 SE +/- 1.04, N = 3 SE +/- 2.63, N = 3 SE +/- 0.64, N = 3 SE +/- 0.64, N = 3 SE +/- 0.08, N = 3 SE +/- 1.59, N = 3 SE +/- 0.78, N = 3 SE +/- 0.23, N = 3 SE +/- 0.17, N = 3 SE +/- 0.26, N = 3 SE +/- 1.53, N = 3 SE +/- 0.09, N = 3 SE +/- 0.43, N = 3 SE +/- 0.64, N = 3 SE +/- 0.48, N = 3 SE +/- 0.35, N = 3 SE +/- 0.19, N = 3 SE +/- 0.04, N = 3 SE +/- 0.40, N = 3 248.68 195.45 191.56 191.11 188.93 180.23 161.50 156.61 153.54 144.64 140.03 137.19 136.69 135.98 131.54 129.65 128.98 110.17 109.46 99.22 60.39 58.58 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: GPT-2 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 EPYC 9684X EPYC 9654 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon 6780E EPYC 9754 Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6766E EPYC 8534PN EPYC 8534P EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 4 8 12 16 20 SE +/- 0.00434, N = 3 SE +/- 0.02114, N = 3 SE +/- 0.01428, N = 3 SE +/- 0.02823, N = 3 SE +/- 0.07483, N = 3 SE +/- 0.01980, N = 3 SE +/- 0.02436, N = 3 SE +/- 0.00351, N = 3 SE +/- 0.06784, N = 3 SE +/- 0.03717, N = 3 SE +/- 0.01181, N = 3 SE +/- 0.00921, N = 3 SE +/- 0.01417, N = 3 SE +/- 0.08283, N = 3 SE +/- 0.00515, N = 3 SE +/- 0.02580, N = 3 SE +/- 0.03809, N = 3 SE +/- 0.03939, N = 3 SE +/- 0.02885, N = 3 SE +/- 0.01997, N = 3 SE +/- 0.01174, N = 3 SE +/- 0.11771, N = 3 4.01687 5.10948 5.21520 5.22781 5.29027 5.54379 6.18478 6.37780 6.50747 6.90619 7.13535 7.28087 7.30822 7.34848 7.59589 7.70511 7.74521 9.06607 9.12452 10.06660 16.54180 17.05530 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
OpenVINO This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Face Detection Retail FP16-INT8 - Device: CPU Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Max 9468 EPYC 9684X 2P EPYC 9684X Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 EPYC 9654 Xeon Platinum 8380 2P EPYC 9654 2P Xeon Max 9480 2P EPYC 9754 Xeon Platinum 8490H Xeon Platinum 8380 EPYC 9754 2P EPYC 8534P EPYC 8534PN Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 5 10 15 20 25 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.02, N = 3 5.18 5.19 5.27 5.34 5.41 5.55 5.72 5.77 5.85 5.93 5.98 6.02 6.25 6.43 6.56 6.60 6.77 7.60 16.41 16.98 18.20 19.23 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenVINO 2024.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.94 / Avg: 130.25 / Max: 140.67 Min: 18.15 / Avg: 159.45 / Max: 171.38 Min: 43.67 / Avg: 229.66 / Max: 249.94 Min: 49.75 / Avg: 255.45 / Max: 289.44 Min: 43.59 / Avg: 299.29 / Max: 327.76 Min: 21 / Avg: 311.43 / Max: 339.91 Min: 31.29 / Avg: 316.48 / Max: 340.73 Min: 35.16 / Avg: 326.91 / Max: 374.78 Min: 58.3 / Avg: 329.39 / Max: 378.11 Min: 72.13 / Avg: 330.52 / Max: 368.64 Min: 77.72 / Avg: 331.34 / Max: 373.94 Min: 30.34 / Avg: 377.27 / Max: 406.75 Min: 105.68 / Avg: 456.22 / Max: 521.95 Min: 100.5 / Avg: 510.02 / Max: 582.06 Min: 108.98 / Avg: 585.34 / Max: 658.24 Min: 39.66 / Avg: 604.27 / Max: 668.8 Min: 59.24 / Avg: 612.41 / Max: 670.98 Min: 83.99 / Avg: 648.24 / Max: 768.1 Min: 165.34 / Avg: 658.39 / Max: 751.53 Min: 203.86 / Avg: 663.32 / Max: 752.94 Min: 196.21 / Avg: 665.03 / Max: 748.66 Min: 65.72 / Avg: 754.92 / Max: 818.94
OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Face Detection Retail FP16-INT8 - Device: CPU EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9754 Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X Xeon 6780E 2P EPYC 9654 Xeon 6766E 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ Xeon Max 9480 EPYC 8534P Xeon Max 9468 Xeon Platinum 8490H Xeon 6780E EPYC 8534PN Xeon 6766E Xeon Platinum 8380 8K 16K 24K 32K 40K SE +/- 10.72, N = 3 SE +/- 15.46, N = 3 SE +/- 46.87, N = 3 SE +/- 17.69, N = 3 SE +/- 12.64, N = 3 SE +/- 8.57, N = 3 SE +/- 3.89, N = 3 SE +/- 12.64, N = 3 SE +/- 7.70, N = 3 SE +/- 41.98, N = 3 SE +/- 2.59, N = 3 SE +/- 30.29, N = 3 SE +/- 16.45, N = 3 SE +/- 4.30, N = 3 SE +/- 7.39, N = 3 SE +/- 2.45, N = 3 SE +/- 18.13, N = 3 SE +/- 5.45, N = 3 SE +/- 0.67, N = 3 SE +/- 7.49, N = 3 SE +/- 2.54, N = 3 SE +/- 15.90, N = 3 38546.66 35809.59 31974.88 24676.86 20937.58 19917.54 18561.91 17252.27 17230.35 16394.06 16038.90 14551.12 13467.28 12317.94 9686.39 9256.11 9087.12 9004.23 8536.39 8279.63 7721.00 5913.04 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: yolov4 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9684X EPYC 9654 EPYC 8534P EPYC 9754 EPYC 8534PN Xeon 6780E Xeon 6766E EPYC 9654 2P EPYC 9684X 2P EPYC 9754 2P Xeon 6766E 2P Xeon 6780E 2P 50 100 150 200 250 SE +/- 0.30, N = 3 SE +/- 0.05, N = 3 SE +/- 0.36, N = 3 SE +/- 0.01, N = 3 SE +/- 0.22, N = 3 SE +/- 0.46, N = 3 SE +/- 0.57, N = 3 SE +/- 0.74, N = 3 SE +/- 0.09, N = 3 SE +/- 0.13, N = 3 SE +/- 0.05, N = 3 SE +/- 0.20, N = 3 SE +/- 0.01, N = 3 SE +/- 0.20, N = 3 SE +/- 0.53, N = 3 SE +/- 0.52, N = 3 SE +/- 0.91, N = 3 SE +/- 0.61, N = 3 SE +/- 0.15, N = 3 SE +/- 2.38, N = 3 SE +/- 2.49, N = 3 SE +/- 1.41, N = 3 47.90 55.12 56.58 61.01 61.83 70.34 70.44 76.42 78.40 79.38 102.04 102.19 114.17 115.79 118.02 129.94 140.14 157.51 158.27 179.99 217.93 219.17 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P 110 220 330 440 550 Min: 18.05 / Avg: 113.49 / Max: 122.89 Min: 19.76 / Avg: 124.93 / Max: 135 Min: 43.74 / Avg: 152.65 / Max: 163.54 Min: 43.53 / Avg: 179.64 / Max: 193.29 Min: 21.66 / Avg: 192.93 / Max: 209.55 Min: 29.58 / Avg: 236.8 / Max: 255.65 Min: 53.1 / Avg: 243.23 / Max: 261.05 Min: 33.83 / Avg: 279.27 / Max: 303.77 Min: 104.75 / Avg: 350.62 / Max: 384.9 Min: 108.42 / Avg: 392.56 / Max: 431.32 Min: 39.89 / Avg: 406.34 / Max: 438.62 Min: 103.81 / Avg: 450.11 / Max: 482.7 Min: 59.75 / Avg: 482.18 / Max: 521.5 Min: 67.07 / Avg: 552.04 / Max: 595.99
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: yolov4 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9684X EPYC 9654 EPYC 8534P EPYC 9754 EPYC 8534PN Xeon 6780E Xeon 6766E EPYC 9654 2P EPYC 9684X 2P EPYC 9754 2P Xeon 6766E 2P Xeon 6780E 2P 5 10 15 20 25 SE +/- 0.13074, N = 3 SE +/- 0.01672, N = 3 SE +/- 0.11211, N = 3 SE +/- 0.00279, N = 3 SE +/- 0.05802, N = 3 SE +/- 0.09305, N = 3 SE +/- 0.11392, N = 3 SE +/- 0.12795, N = 3 SE +/- 0.01439, N = 3 SE +/- 0.02010, N = 3 SE +/- 0.00491, N = 3 SE +/- 0.01934, N = 3 SE +/- 0.00112, N = 3 SE +/- 0.01474, N = 3 SE +/- 0.03807, N = 3 SE +/- 0.03075, N = 3 SE +/- 0.04609, N = 3 SE +/- 0.02467, N = 3 SE +/- 0.00606, N = 3 SE +/- 0.07368, N = 3 SE +/- 0.05214, N = 3 SE +/- 0.02942, N = 3 20.87950 18.14130 17.67470 16.38980 16.17420 14.21690 14.19800 13.08690 12.75460 12.59740 9.80008 9.78516 8.75814 8.63594 8.47283 7.69583 7.13587 6.34891 6.31828 5.55767 4.58960 4.56278 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: yolov4 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9684X EPYC 9654 EPYC 8534P EPYC 9754 EPYC 8534PN Xeon 6780E Xeon 6766E EPYC 9654 2P EPYC 9684X 2P EPYC 9754 2P Xeon 6766E 2P Xeon 6780E 2P 50 100 150 200 250 SE +/- 0.30, N = 3 SE +/- 0.05, N = 3 SE +/- 0.36, N = 3 SE +/- 0.01, N = 3 SE +/- 0.22, N = 3 SE +/- 0.46, N = 3 SE +/- 0.57, N = 3 SE +/- 0.74, N = 3 SE +/- 0.09, N = 3 SE +/- 0.13, N = 3 SE +/- 0.05, N = 3 SE +/- 0.20, N = 3 SE +/- 0.01, N = 3 SE +/- 0.20, N = 3 SE +/- 0.53, N = 3 SE +/- 0.52, N = 3 SE +/- 0.91, N = 3 SE +/- 0.61, N = 3 SE +/- 0.15, N = 3 SE +/- 2.38, N = 3 SE +/- 2.49, N = 3 SE +/- 1.41, N = 3 47.90 55.12 56.58 61.01 61.83 70.34 70.44 76.42 78.40 79.38 102.04 102.19 114.17 115.79 118.02 129.94 140.14 157.51 158.27 179.99 217.93 219.17 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
GraphicsMagick This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.43 Operation: Noise-Gaussian Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ EPYC 9684X EPYC 9754 2P Xeon Max 9480 2P EPYC 9654 Xeon Max 9480 Xeon Platinum 8490H Xeon 6780E Xeon Platinum 8380 2P EPYC 9754 Xeon Max 9468 2P Xeon Max 9468 Xeon 6766E Xeon 6780E 2P EPYC 8534P Xeon Platinum 8380 EPYC 8534PN Xeon 6766E 2P 40 80 120 160 200 SE +/- 0.88, N = 3 SE +/- 1.15, N = 3 SE +/- 0.67, N = 3 SE +/- 0.67, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.58, N = 3 SE +/- 0.33, N = 3 SE +/- 0.58, N = 3 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 SE +/- 0.58, N = 3 SE +/- 0.00, N = 3 SE +/- 0.58, N = 3 SE +/- 0.58, N = 3 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 SE +/- 1.44, N = 4 SE +/- 0.33, N = 3 SE +/- 0.67, N = 3 SE +/- 0.00, N = 3 SE +/- 0.33, N = 3 194 182 179 178 171 165 164 164 160 158 158 155 153 151 150 150 145 126 123 121 120 118 1. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp Iterations Per Minute Per Watt
OpenBenchmarking.org Iterations Per Minute Per Watt, More Is Better GraphicsMagick 1.3.43 Operation: Noise-Gaussian Xeon 6766E EPYC 8534PN EPYC 8534P Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8490H EPYC 9754 2P Xeon Platinum 8380 Xeon Max 9480 EPYC 9654 2P Xeon Max 9468 Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.3443 0.6886 1.0329 1.3772 1.7215 1.530 1.507 1.420 1.417 1.226 1.041 0.932 0.778 0.740 0.729 0.716 0.674 0.620 0.614 0.597 0.552 0.544 0.540 0.482 0.467 0.345 0.308 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better GraphicsMagick 1.3.43 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon 6766E 2P Xeon 6780E 2P EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9480 Xeon Max 9468 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 90 180 270 360 450 Min: 16.29 / Avg: 79.6 / Max: 84.69 Min: 17.9 / Avg: 86.65 / Max: 95.02 Min: 43.66 / Avg: 94.78 / Max: 111.52 Min: 43.47 / Avg: 109.4 / Max: 130.3 Min: 21.53 / Avg: 123.19 / Max: 140.7 Min: 29.47 / Avg: 153.63 / Max: 171.14 Min: 104.1 / Avg: 159.54 / Max: 190.88 Min: 98.17 / Avg: 172.89 / Max: 203.08 Min: 33.85 / Avg: 176.95 / Max: 194.06 Min: 53.3 / Avg: 195.01 / Max: 207.9 Min: 34.53 / Avg: 219.79 / Max: 241.79 Min: 59.26 / Avg: 220.76 / Max: 250.24 Min: 39.77 / Avg: 243.28 / Max: 272.18 Min: 75.67 / Avg: 257.27 / Max: 283.18 Min: 74.91 / Avg: 271.55 / Max: 302.98 Min: 59.18 / Avg: 297.92 / Max: 323.55 Min: 99.29 / Avg: 317.68 / Max: 360.97 Min: 66.05 / Avg: 331.54 / Max: 359.49 Min: 82.86 / Avg: 356.85 / Max: 403.74 Min: 160.72 / Avg: 389.84 / Max: 423.25 Min: 196.74 / Avg: 476 / Max: 507.38 Min: 203.35 / Avg: 486.24 / Max: 527.71
GPAW GPAW is a density-functional theory (DFT) Python code based on the projector-augmented wave (PAW) method and the atomic simulation environment (ASE). Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better GPAW 23.6 Input: Carbon Nanotube EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P EPYC 9684X EPYC 9754 EPYC 9654 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Platinum 8592+ Xeon Max 9480 2P Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 20 40 60 80 100 SE +/- 0.30, N = 15 SE +/- 0.29, N = 15 SE +/- 0.18, N = 3 SE +/- 0.11, N = 3 SE +/- 0.14, N = 3 SE +/- 0.21, N = 3 SE +/- 0.10, N = 3 SE +/- 0.55, N = 3 SE +/- 0.08, N = 3 SE +/- 0.12, N = 3 SE +/- 0.27, N = 3 SE +/- 0.24, N = 3 SE +/- 0.60, N = 3 SE +/- 0.08, N = 3 SE +/- 0.03, N = 3 SE +/- 0.07, N = 3 SE +/- 0.01, N = 3 SE +/- 0.12, N = 3 24.13 25.84 27.15 30.00 33.12 34.87 38.05 41.56 42.87 43.05 43.91 47.78 48.45 48.49 54.38 63.80 66.71 76.18 1. (CC) gcc options: -shared -lxc -lblas -lmpi CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better GPAW 23.6 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 18.16 / Avg: 136.93 / Max: 153.06 Min: 18.71 / Avg: 154.54 / Max: 175.1 Min: 53.47 / Avg: 257.77 / Max: 289.43 Min: 23.61 / Avg: 293.6 / Max: 347.81 Min: 28.47 / Avg: 300.58 / Max: 350.83 Min: 24.08 / Avg: 314.73 / Max: 374.48 Min: 59.13 / Avg: 322.66 / Max: 373.01 Min: 75.93 / Avg: 324.94 / Max: 375.46 Min: 72.36 / Avg: 327.19 / Max: 373.7 Min: 34.27 / Avg: 353.01 / Max: 409.6 Min: 103.43 / Avg: 497.52 / Max: 576.52 Min: 43.78 / Avg: 513.29 / Max: 700.14 Min: 57.4 / Avg: 562.66 / Max: 711.92 Min: 80.11 / Avg: 568.78 / Max: 750.58 Min: 153.52 / Avg: 606.84 / Max: 749.04 Min: 203.73 / Avg: 621.08 / Max: 746.35 Min: 213.42 / Avg: 622.05 / Max: 746.91 Min: 65.21 / Avg: 641.29 / Max: 819.6
GraphicsMagick This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.43 Operation: Sharpen Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon 6780E Xeon Platinum 8490H 2P EPYC 9654 2P Xeon 6766E Xeon Max 9480 2P EPYC 9754 EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 80 160 240 320 400 SE +/- 0.67, N = 3 SE +/- 0.88, N = 3 SE +/- 3.00, N = 3 SE +/- 0.88, N = 3 SE +/- 1.67, N = 3 SE +/- 0.00, N = 3 SE +/- 1.67, N = 3 SE +/- 1.20, N = 3 SE +/- 0.00, N = 3 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 SE +/- 0.00, N = 3 SE +/- 1.20, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 381 336 310 297 284 281 265 262 253 227 207 201 199 173 167 157 144 130 114 112 105 79 1. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp Iterations Per Minute Per Watt
OpenBenchmarking.org Iterations Per Minute Per Watt, More Is Better GraphicsMagick 1.3.43 Operation: Sharpen Xeon 6766E Xeon 6780E Xeon 6766E 2P EPYC 8534PN Xeon 6780E 2P EPYC 8534P EPYC 9754 EPYC 9754 2P EPYC 9654 EPYC 9684X EPYC 9654 2P Xeon Platinum 8592+ EPYC 9684X 2P Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9468 Xeon Max 9480 2P Xeon Platinum 8380 Xeon Max 9468 2P Xeon Platinum 8380 2P 0.2781 0.5562 0.8343 1.1124 1.3905 1.236 1.074 0.965 0.904 0.894 0.881 0.838 0.690 0.674 0.589 0.544 0.512 0.492 0.476 0.453 0.419 0.405 0.345 0.343 0.310 0.301 0.285 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better GraphicsMagick 1.3.43 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon 6780E Xeon Max 9480 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 120 240 360 480 600 Min: 15.82 / Avg: 116.15 / Max: 127.18 Min: 17.43 / Avg: 127.19 / Max: 145.11 Min: 43.7 / Avg: 204.72 / Max: 219.13 Min: 22 / Avg: 247.13 / Max: 273.84 Min: 53.14 / Avg: 254.91 / Max: 285.61 Min: 28.8 / Avg: 256.57 / Max: 278.61 Min: 43.66 / Avg: 261.59 / Max: 286.06 Min: 74.28 / Avg: 321.22 / Max: 366.4 Min: 26.5 / Avg: 326.08 / Max: 362.61 Min: 62.29 / Avg: 329.54 / Max: 363.35 Min: 74.41 / Avg: 330.11 / Max: 353.01 Min: 34.61 / Avg: 341.04 / Max: 370.17 Min: 104.82 / Avg: 348.18 / Max: 384.37 Min: 107.76 / Avg: 425.97 / Max: 469.34 Min: 38.07 / Avg: 448.99 / Max: 494.64 Min: 59.05 / Avg: 481.76 / Max: 531.74 Min: 102.73 / Avg: 504.87 / Max: 544.93 Min: 65.27 / Avg: 603.16 / Max: 655.4 Min: 74.46 / Avg: 627.19 / Max: 683.79 Min: 148.52 / Avg: 632.93 / Max: 684.15 Min: 197.85 / Avg: 660.39 / Max: 706.79 Min: 206.23 / Avg: 662.12 / Max: 708.79
Timed Linux Kernel Compilation This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Timed Linux Kernel Compilation 6.8 Build: defconfig Xeon Platinum 8592+ 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9754 2P Xeon Max 9468 2P Xeon 6780E 2P EPYC 9684X EPYC 9654 Xeon Platinum 8592+ Xeon 6766E 2P Xeon 6780E EPYC 9754 Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon 6766E Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 13 26 39 52 65 SE +/- 0.23, N = 7 SE +/- 0.17, N = 13 SE +/- 0.33, N = 3 SE +/- 0.43, N = 3 SE +/- 0.31, N = 5 SE +/- 0.28, N = 7 SE +/- 0.40, N = 4 SE +/- 0.28, N = 8 SE +/- 0.46, N = 3 SE +/- 0.43, N = 3 SE +/- 0.39, N = 4 SE +/- 0.22, N = 14 SE +/- 0.37, N = 6 SE +/- 0.44, N = 3 SE +/- 0.42, N = 5 SE +/- 0.55, N = 3 SE +/- 0.57, N = 3 SE +/- 0.39, N = 6 SE +/- 0.59, N = 3 SE +/- 0.50, N = 4 SE +/- 0.54, N = 4 SE +/- 0.33, N = 3 25.87 25.87 27.68 29.98 30.53 30.68 32.14 32.66 32.89 34.57 34.59 35.55 37.64 38.53 39.41 39.45 40.50 41.10 43.42 44.28 46.35 55.74 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Timed Linux Kernel Compilation 6.8 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon 6766E 2P Xeon Platinum 8380 Xeon 6780E 2P EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 14.06 / Avg: 87.28 / Max: 138.54 Min: 16.1 / Avg: 97.22 / Max: 163.42 Min: 43.51 / Avg: 106.04 / Max: 250.64 Min: 21.97 / Avg: 123.43 / Max: 312.29 Min: 20.78 / Avg: 155.1 / Max: 328.08 Min: 27.91 / Avg: 175.53 / Max: 322.12 Min: 104.46 / Avg: 183.63 / Max: 516.23 Min: 48.64 / Avg: 197.71 / Max: 290.64 Min: 106.91 / Avg: 202.81 / Max: 628.9 Min: 32.62 / Avg: 210.93 / Max: 409.9 Min: 26.35 / Avg: 214.95 / Max: 377.53 Min: 60.49 / Avg: 234.11 / Max: 377.45 Min: 69.92 / Avg: 260.34 / Max: 376.76 Min: 73.73 / Avg: 265.23 / Max: 376.52 Min: 38.17 / Avg: 283.46 / Max: 659.24 Min: 52.83 / Avg: 317.6 / Max: 662.66 Min: 98.92 / Avg: 340.2 / Max: 582.62 Min: 62.86 / Avg: 375.56 / Max: 824.56 Min: 73.91 / Avg: 384.63 / Max: 756.99 Min: 150.36 / Avg: 425.97 / Max: 755.88 Min: 176.87 / Avg: 486.28 / Max: 754.8 Min: 189.61 / Avg: 492.26 / Max: 752.95
GraphicsMagick This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.43 Operation: Enhanced EPYC 9684X 2P EPYC 9754 2P Xeon 6780E 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon 6766E 2P Xeon Platinum 8490H 2P EPYC 9754 Xeon Max 9480 2P EPYC 9684X Xeon 6780E Xeon 6766E Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon Max 9480 EPYC 8534P EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 90 180 270 360 450 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 SE +/- 1.53, N = 3 SE +/- 1.15, N = 3 SE +/- 1.20, N = 3 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 SE +/- 1.00, N = 3 SE +/- 0.67, N = 3 SE +/- 0.33, N = 3 SE +/- 0.00, N = 3 SE +/- 2.60, N = 3 SE +/- 0.33, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 1.45, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 408 403 389 374 368 350 342 304 301 300 299 270 269 264 228 216 215 181 172 163 160 123 1. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp Iterations Per Minute Per Watt
OpenBenchmarking.org Iterations Per Minute Per Watt, More Is Better GraphicsMagick 1.3.43 Operation: Enhanced EPYC 8534PN Xeon 6766E EPYC 8534P EPYC 9754 Xeon 6780E Xeon 6766E 2P EPYC 9654 EPYC 9754 2P Xeon 6780E 2P EPYC 9684X EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 2P Xeon Platinum 8380 2P Xeon Max 9468 2P 0.3308 0.6616 0.9924 1.3232 1.654 1.470 1.391 1.348 1.342 1.211 1.086 1.067 1.014 0.999 0.935 0.815 0.742 0.696 0.656 0.614 0.568 0.549 0.484 0.484 0.466 0.435 0.410 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better GraphicsMagick 1.3.43 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Max 9468 2P 120 240 360 480 600 Min: 16.93 / Avg: 110.88 / Max: 122.16 Min: 18.4 / Avg: 127.55 / Max: 140 Min: 43.58 / Avg: 194.07 / Max: 209.22 Min: 20.16 / Avg: 226.47 / Max: 243.63 Min: 43.55 / Avg: 246.9 / Max: 265.73 Min: 26.64 / Avg: 247.53 / Max: 267.87 Min: 52.22 / Avg: 254.21 / Max: 282.55 Min: 33.82 / Avg: 320.71 / Max: 343.53 Min: 104.13 / Avg: 322.39 / Max: 344.94 Min: 28.79 / Avg: 327.64 / Max: 359.97 Min: 61.03 / Avg: 329.43 / Max: 359.83 Min: 75.94 / Avg: 329.68 / Max: 362.6 Min: 73.74 / Avg: 330.74 / Max: 355.56 Min: 105.4 / Avg: 389.46 / Max: 429.53 Min: 40.42 / Avg: 397.41 / Max: 433.85 Min: 59.43 / Avg: 451.41 / Max: 489.55 Min: 101.45 / Avg: 493.87 / Max: 534.32 Min: 64.84 / Avg: 549.68 / Max: 597.34 Min: 161.82 / Avg: 601.76 / Max: 650.47 Min: 74.15 / Avg: 608.66 / Max: 666.76 Min: 202.79 / Avg: 645.53 / Max: 683.28 Min: 198.68 / Avg: 656.05 / Max: 702.16
ONNX Runtime ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: super-resolution-10 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9654 EPYC 9684X EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 4 8 12 16 20 SE +/- 0.03569, N = 3 SE +/- 0.02988, N = 3 SE +/- 0.05586, N = 3 SE +/- 0.02534, N = 3 SE +/- 0.01694, N = 3 SE +/- 0.02062, N = 3 SE +/- 0.02822, N = 3 SE +/- 0.01290, N = 3 SE +/- 0.03854, N = 3 SE +/- 0.02155, N = 3 SE +/- 0.03015, N = 3 SE +/- 0.08338, N = 3 SE +/- 0.05765, N = 3 SE +/- 0.07856, N = 3 SE +/- 0.02898, N = 3 SE +/- 0.05412, N = 3 SE +/- 0.06993, N = 3 SE +/- 0.13507, N = 3 SE +/- 0.00657, N = 3 SE +/- 0.00867, N = 3 SE +/- 0.02665, N = 3 SE +/- 0.06373, N = 3 4.26352 4.65431 4.86604 5.10914 5.26075 5.29287 5.59763 5.67175 5.74921 5.88628 9.08955 9.18524 10.30260 10.39770 10.55370 11.45800 11.50240 12.80930 13.17890 14.47610 15.27840 16.58280 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ONNX Runtime 1.17 CPU Power Consumption Monitor Xeon 6766E EPYC 8534PN EPYC 8534P Xeon 6780E EPYC 9754 EPYC 9654 Xeon 6766E 2P Xeon 6780E 2P EPYC 9684X EPYC 9754 2P Xeon Platinum 8380 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P 60 120 180 240 300 Min: 43.65 / Avg: 75.71 / Max: 78.31 Min: 16.21 / Avg: 80.29 / Max: 85.38 Min: 17.58 / Avg: 85.08 / Max: 90.86 Min: 43.53 / Avg: 86.18 / Max: 89.75 Min: 20.2 / Avg: 107.51 / Max: 113.85 Min: 29.43 / Avg: 145.22 / Max: 154.34 Min: 103.7 / Avg: 151.35 / Max: 165.49 Min: 84.94 / Avg: 161.97 / Max: 179.97 Min: 34.49 / Avg: 165.58 / Max: 176.42 Min: 40.67 / Avg: 218.35 / Max: 232.79 Min: 50.48 / Avg: 227.45 / Max: 243.02 Min: 58.04 / Avg: 272.09 / Max: 288.97 Min: 64.69 / Avg: 300.88 / Max: 323.81 Min: 102.14 / Avg: 339.94 / Max: 362.78
Result
OpenBenchmarking.org Inferences Per Second, More Is Better ONNX Runtime 1.17 Model: super-resolution-10 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9654 EPYC 9684X EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 50 100 150 200 250 SE +/- 1.98, N = 3 SE +/- 1.38, N = 3 SE +/- 2.34, N = 3 SE +/- 0.98, N = 3 SE +/- 0.61, N = 3 SE +/- 0.74, N = 3 SE +/- 0.90, N = 3 SE +/- 0.40, N = 3 SE +/- 1.17, N = 3 SE +/- 0.62, N = 3 SE +/- 0.36, N = 3 SE +/- 0.99, N = 3 SE +/- 0.54, N = 3 SE +/- 0.72, N = 3 SE +/- 0.26, N = 3 SE +/- 0.41, N = 3 SE +/- 0.53, N = 3 SE +/- 0.83, N = 3 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.11, N = 3 SE +/- 0.23, N = 3 234.54 214.81 205.53 195.71 190.06 188.89 178.62 176.29 173.91 169.85 110.00 108.87 97.05 96.17 94.74 87.26 86.92 78.07 75.87 69.07 65.43 60.29 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt Inference Time Cost (ms)
OpenBenchmarking.org Inference Time Cost (ms), Fewer Is Better ONNX Runtime 1.17 Model: super-resolution-10 - Device: CPU - Executor: Parallel Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9654 EPYC 9684X EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 4 8 12 16 20 SE +/- 0.03569, N = 3 SE +/- 0.02988, N = 3 SE +/- 0.05586, N = 3 SE +/- 0.02534, N = 3 SE +/- 0.01694, N = 3 SE +/- 0.02062, N = 3 SE +/- 0.02822, N = 3 SE +/- 0.01290, N = 3 SE +/- 0.03854, N = 3 SE +/- 0.02155, N = 3 SE +/- 0.03015, N = 3 SE +/- 0.08338, N = 3 SE +/- 0.05765, N = 3 SE +/- 0.07856, N = 3 SE +/- 0.02898, N = 3 SE +/- 0.05412, N = 3 SE +/- 0.06993, N = 3 SE +/- 0.13507, N = 3 SE +/- 0.00657, N = 3 SE +/- 0.00867, N = 3 SE +/- 0.02665, N = 3 SE +/- 0.06373, N = 3 4.26352 4.65431 4.86604 5.10914 5.26075 5.29287 5.59763 5.67175 5.74921 5.88628 9.08955 9.18524 10.30260 10.39770 10.55370 11.45800 11.50240 12.80930 13.17890 14.47610 15.27840 16.58280 1. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt
GraphicsMagick This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.43 Operation: Swirl Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P EPYC 9754 2P Xeon 6780E Xeon Max 9480 2P EPYC 9654 2P EPYC 9684X EPYC 9754 Xeon Platinum 8592+ Xeon 6780E 2P EPYC 9654 Xeon Max 9468 2P Xeon 6766E Xeon Platinum 8490H Xeon 6766E 2P Xeon Max 9480 Xeon Platinum 8380 2P Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 150 300 450 600 750 SE +/- 0.58, N = 3 SE +/- 1.76, N = 3 SE +/- 0.67, N = 3 SE +/- 0.33, N = 3 SE +/- 1.15, N = 3 SE +/- 0.67, N = 3 SE +/- 0.33, N = 3 SE +/- 0.00, N = 3 SE +/- 0.33, N = 3 SE +/- 0.88, N = 3 SE +/- 3.00, N = 3 SE +/- 0.33, N = 3 SE +/- 1.33, N = 3 SE +/- 4.91, N = 3 SE +/- 1.15, N = 3 SE +/- 2.33, N = 3 SE +/- 0.67, N = 3 SE +/- 4.41, N = 3 SE +/- 0.58, N = 3 SE +/- 0.33, N = 3 SE +/- 0.88, N = 3 SE +/- 0.33, N = 3 689 617 583 567 566 558 553 546 535 529 517 511 509 504 502 474 432 403 394 379 363 278 1. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp Iterations Per Minute Per Watt
OpenBenchmarking.org Iterations Per Minute Per Watt, More Is Better GraphicsMagick 1.3.43 Operation: Swirl EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8592+ EPYC 9754 2P Xeon Platinum 8490H EPYC 9654 2P Xeon Max 9480 Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Max 9480 2P Xeon Platinum 8380 2P Xeon Max 9468 2P 0.7398 1.4796 2.2194 2.9592 3.699 3.288 3.279 3.093 2.944 2.779 2.290 2.036 1.923 1.872 1.724 1.710 1.617 1.445 1.341 1.307 1.298 1.193 1.137 1.097 0.964 0.943 0.851 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better GraphicsMagick 1.3.43 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon 6766E 2P Xeon Platinum 8380 EPYC 9684X Xeon 6780E 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 110 220 330 440 550 Min: 17.2 / Avg: 110.39 / Max: 118.71 Min: 18.54 / Avg: 122.54 / Max: 132.44 Min: 43.71 / Avg: 153.73 / Max: 166.75 Min: 43.48 / Avg: 192.24 / Max: 208.39 Min: 21.51 / Avg: 192.51 / Max: 206.12 Min: 30.38 / Avg: 223.17 / Max: 243.96 Min: 104.8 / Avg: 246.46 / Max: 265.1 Min: 52.31 / Avg: 253.3 / Max: 271.57 Min: 34.46 / Avg: 268.23 / Max: 288.18 Min: 107.08 / Avg: 276.13 / Max: 303.23 Min: 25.3 / Avg: 306.84 / Max: 336.43 Min: 59.04 / Avg: 310.53 / Max: 334.67 Min: 70.83 / Avg: 322.19 / Max: 348.63 Min: 72.86 / Avg: 330.14 / Max: 351.19 Min: 40.54 / Avg: 331.64 / Max: 362.33 Min: 56.5 / Avg: 382.8 / Max: 426.76 Min: 104.89 / Avg: 427.21 / Max: 467.28 Min: 62.97 / Avg: 449.12 / Max: 484.89 Min: 80.88 / Avg: 527.29 / Max: 574.8 Min: 159.26 / Avg: 542.66 / Max: 585.44 Min: 192.72 / Avg: 578.67 / Max: 624.15 Min: 191.82 / Avg: 598.34 / Max: 636.3
RocksDB This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Op/s, More Is Better RocksDB 9.0 Test: Random Read Xeon 6780E 2P EPYC 9754 2P Xeon 6766E 2P EPYC 9684X 2P EPYC 9654 2P Xeon 6780E Xeon Platinum 8592+ 2P Xeon 6766E EPYC 9754 Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9480 2P EPYC 9654 Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 8534P Xeon Platinum 8490H Xeon Platinum 8380 2P EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 300M 600M 900M 1200M 1500M SE +/- 1897433.73, N = 3 SE +/- 1154820.73, N = 3 SE +/- 4448269.22, N = 3 SE +/- 2927744.32, N = 3 SE +/- 599126.55, N = 3 SE +/- 2130676.22, N = 3 SE +/- 82996.27, N = 3 SE +/- 1336933.52, N = 3 SE +/- 153963.59, N = 3 SE +/- 1411218.60, N = 3 SE +/- 326038.01, N = 3 SE +/- 785893.13, N = 3 SE +/- 610621.13, N = 3 SE +/- 1120857.00, N = 3 SE +/- 1119855.47, N = 3 SE +/- 206014.27, N = 3 SE +/- 3338569.32, N = 3 SE +/- 704022.38, N = 3 SE +/- 211259.57, N = 3 SE +/- 86315.43, N = 3 SE +/- 81327.78, N = 3 SE +/- 744772.41, N = 3 1272907561 1135206564 1115720742 1031209151 815145110 645309483 578392800 574749552 573235452 525049998 523310091 473755946 409272949 405533173 297517946 266186537 265366170 251234103 234992636 234469939 205346154 125584953 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread Op/s Per Watt
OpenBenchmarking.org Op/s Per Watt, More Is Better RocksDB 9.0 Test: Random Read Xeon 6766E Xeon 6766E 2P Xeon 6780E Xeon 6780E 2P EPYC 8534PN EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 9654 EPYC 9654 2P EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9480 Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 500K 1000K 1500K 2000K 2500K 2436056.02 2359232.60 2069916.27 2038972.60 1972313.57 1957165.94 1907427.82 1780768.80 1523041.95 1477897.34 1362651.04 1341710.00 904750.01 878232.54 802721.25 791368.94 709419.37 704496.50 663946.02 608125.97 490897.23 489869.48 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better RocksDB 9.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 EPYC 9654 EPYC 9754 Xeon Max 9468 Xeon 6780E Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 18.15 / Avg: 119.15 / Max: 127.37 Min: 19.07 / Avg: 149.48 / Max: 159.86 Min: 43.66 / Avg: 235.93 / Max: 270.68 Min: 53.91 / Avg: 255.83 / Max: 291.63 Min: 24.72 / Avg: 268.72 / Max: 291.05 Min: 21.45 / Avg: 292.89 / Max: 315.65 Min: 73.12 / Avg: 309.28 / Max: 327.48 Min: 43.52 / Avg: 311.76 / Max: 357.69 Min: 25.38 / Avg: 328.84 / Max: 377.98 Min: 54.21 / Avg: 330.58 / Max: 378.42 Min: 77.85 / Avg: 332.82 / Max: 365.47 Min: 33.04 / Avg: 384.04 / Max: 411.06 Min: 104.97 / Avg: 472.92 / Max: 541.33 Min: 100.64 / Avg: 512.86 / Max: 581.28 Min: 53.06 / Avg: 551.56 / Max: 593.87 Min: 40.42 / Avg: 595.15 / Max: 641.39 Min: 107.45 / Avg: 624.29 / Max: 712.89 Min: 63.13 / Avg: 658.59 / Max: 754.12 Min: 145.71 / Avg: 663.47 / Max: 754.3 Min: 186.94 / Avg: 666.86 / Max: 752.6 Min: 195.53 / Avg: 667.81 / Max: 755.48 Min: 61.31 / Avg: 768.58 / Max: 822.92
miniBUDE MiniBUDE is a mini application for the the core computation of the Bristol University Docking Engine (BUDE). This test profile currently makes use of the OpenMP implementation of miniBUDE for CPU benchmarking. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Billion Interactions/s, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM2 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P EPYC 9754 Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 Xeon 6780E Xeon 6766E 70 140 210 280 350 SE +/- 2.28, N = 3 SE +/- 0.08, N = 3 SE +/- 3.85, N = 3 SE +/- 0.01, N = 3 SE +/- 1.93, N = 15 SE +/- 1.71, N = 3 SE +/- 0.93, N = 3 SE +/- 0.99, N = 3 SE +/- 0.93, N = 3 SE +/- 1.48, N = 3 SE +/- 0.03, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.75, N = 3 SE +/- 0.31, N = 3 SE +/- 0.08, N = 3 SE +/- 0.03, N = 3 SE +/- 0.96, N = 4 SE +/- 0.31, N = 3 SE +/- 0.11, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 309.22 291.28 288.54 232.75 191.82 187.68 171.77 140.83 133.99 127.85 108.12 107.72 97.98 97.08 86.97 86.88 78.24 78.11 70.21 49.20 43.73 39.35 1. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm Billion Interactions/s Per Watt
OpenBenchmarking.org Billion Interactions/s Per Watt, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM2 EPYC 9754 EPYC 8534PN EPYC 8534P EPYC 9754 2P EPYC 9654 EPYC 9654 2P EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H 2P Xeon 6766E Xeon 6766E 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon Platinum 8380 2P Xeon 6780E 2P Xeon Platinum 8380 0.1906 0.3812 0.5718 0.7624 0.953 0.847 0.789 0.746 0.669 0.605 0.588 0.551 0.532 0.334 0.319 0.275 0.248 0.245 0.244 0.225 0.216 0.215 0.211 0.196 0.196 0.195 0.191 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better miniBUDE 20210901 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H Xeon Platinum 8592+ EPYC 9684X Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 16.52 / Avg: 124.23 / Max: 135.64 Min: 18.74 / Avg: 144.45 / Max: 156.43 Min: 43.66 / Avg: 174.56 / Max: 194.55 Min: 43.56 / Avg: 223.09 / Max: 250.29 Min: 52.78 / Avg: 257.94 / Max: 287.95 Min: 20.19 / Avg: 274.94 / Max: 327 Min: 28.02 / Avg: 283.82 / Max: 328.33 Min: 72.43 / Avg: 286.83 / Max: 306.63 Min: 75.23 / Avg: 314.81 / Max: 352.37 Min: 62.05 / Avg: 316.19 / Max: 372.81 Min: 23.61 / Avg: 323.61 / Max: 376.85 Min: 33.63 / Avg: 340.93 / Max: 409.44 Min: 105.16 / Avg: 361.96 / Max: 383.12 Min: 42.31 / Avg: 431.39 / Max: 620.63 Min: 107.71 / Avg: 444.54 / Max: 472.65 Min: 104.71 / Avg: 494.9 / Max: 576.68 Min: 59.15 / Avg: 495.6 / Max: 654.49 Min: 173.4 / Avg: 578.13 / Max: 755.42 Min: 68.45 / Avg: 581.3 / Max: 795.09 Min: 63.99 / Avg: 600.52 / Max: 754.84 Min: 193.86 / Avg: 605.49 / Max: 719.24 Min: 203.67 / Avg: 621.91 / Max: 746.15
OpenBenchmarking.org GFInst/s, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM2 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P EPYC 9754 Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 Xeon 6780E Xeon 6766E 1700 3400 5100 6800 8500 SE +/- 57.10, N = 3 SE +/- 2.03, N = 3 SE +/- 96.25, N = 3 SE +/- 0.20, N = 3 SE +/- 48.20, N = 15 SE +/- 42.85, N = 3 SE +/- 23.15, N = 3 SE +/- 24.63, N = 3 SE +/- 23.14, N = 3 SE +/- 36.96, N = 3 SE +/- 0.70, N = 3 SE +/- 0.13, N = 3 SE +/- 0.10, N = 3 SE +/- 18.66, N = 3 SE +/- 7.77, N = 3 SE +/- 1.96, N = 3 SE +/- 0.82, N = 3 SE +/- 23.93, N = 4 SE +/- 7.81, N = 3 SE +/- 2.69, N = 3 SE +/- 0.09, N = 3 SE +/- 0.03, N = 3 7730.47 7281.96 7213.43 5818.74 4795.46 4692.11 4294.28 3520.71 3349.76 3196.12 2703.09 2692.87 2449.45 2426.88 2174.19 2172.10 1955.88 1952.61 1755.33 1230.08 1093.21 983.71 1. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
Coremark This is a test of EEMBC CoreMark processor benchmark. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Iterations/Sec, More Is Better Coremark 1.0 CoreMark Size 666 - Iterations Per Second EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8592+ 2P EPYC 9754 EPYC 9684X EPYC 9654 Xeon Platinum 8490H 2P Xeon 6780E Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6766E Xeon Platinum 8592+ EPYC 8534P Xeon Platinum 8490H EPYC 8534PN Xeon Max 9480 Xeon Platinum 8380 2P Xeon Max 9468 Xeon Platinum 8380 2M 4M 6M 8M 10M SE +/- 76743.24, N = 3 SE +/- 54771.53, N = 3 SE +/- 2306.54, N = 3 SE +/- 7619.10, N = 3 SE +/- 9906.84, N = 3 SE +/- 9700.21, N = 3 SE +/- 4775.53, N = 3 SE +/- 54326.71, N = 3 SE +/- 10132.66, N = 3 SE +/- 12743.41, N = 3 SE +/- 40214.40, N = 15 SE +/- 20649.43, N = 3 SE +/- 7219.73, N = 3 SE +/- 10294.77, N = 3 SE +/- 4635.95, N = 3 SE +/- 1374.66, N = 3 SE +/- 9256.47, N = 3 SE +/- 11678.22, N = 3 SE +/- 5202.86, N = 3 SE +/- 30163.42, N = 15 SE +/- 1747.99, N = 3 SE +/- 927.64, N = 3 8448083.73 8171241.71 7975907.45 7302247.14 6575205.54 4966250.78 4525288.00 4152864.88 3926265.28 3902076.24 3340274.03 3279262.53 2993136.60 2915609.75 2369998.36 2274636.13 2194630.60 2069432.53 1855135.33 1839444.40 1614763.70 1218453.34 1. (CC) gcc options: -O2 -lrt" -lrt Iterations/Sec Per Watt
OpenBenchmarking.org Iterations/Sec Per Watt, More Is Better Coremark 1.0 CoreMark Size 666 - Iterations Per Second Xeon 6766E 2P Xeon 6780E 2P Xeon 6766E EPYC 8534PN EPYC 9754 EPYC 9754 2P EPYC 8534P Xeon 6780E EPYC 9654 2P EPYC 9684X 2P EPYC 9654 EPYC 9684X Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9468 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 5K 10K 15K 20K 25K 21799.95 18406.92 18400.49 16836.69 16764.47 16699.88 15712.09 15612.90 14883.36 13681.18 13254.14 12144.31 8414.46 7555.56 7477.77 7118.47 6571.62 6163.40 5880.62 5763.20 5016.56 4373.42 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Coremark 1.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 Xeon Max 9468 EPYC 9754 Xeon Max 9480 Xeon Platinum 8490H EPYC 9654 Xeon 6766E 2P Xeon Platinum 8592+ EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9754 2P Xeon Max 9468 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P 140 280 420 560 700 Min: 14.91 / Avg: 122.91 / Max: 138.78 Min: 16.44 / Avg: 144.77 / Max: 164.64 Min: 43.57 / Avg: 158.45 / Max: 217.13 Min: 21.83 / Avg: 213.94 / Max: 282.7 Min: 50.82 / Avg: 242.89 / Max: 290.97 Min: 76.41 / Avg: 261.99 / Max: 288.93 Min: 21.14 / Avg: 269.93 / Max: 353.42 Min: 77.43 / Avg: 282.3 / Max: 314.07 Min: 54.25 / Avg: 293.49 / Max: 341.26 Min: 26.88 / Avg: 296.23 / Max: 351.89 Min: 104.32 / Avg: 301.62 / Max: 446.07 Min: 34.34 / Avg: 313.68 / Max: 378.24 Min: 32.2 / Avg: 341.96 / Max: 411.86 Min: 106.84 / Avg: 396.71 / Max: 563.32 Min: 100.85 / Avg: 420.6 / Max: 582.07 Min: 42.53 / Avg: 489.3 / Max: 675.15 Min: 188.96 / Avg: 519.35 / Max: 613.18 Min: 57.24 / Avg: 535.89 / Max: 711.09 Min: 163.01 / Avg: 548.16 / Max: 709.21 Min: 206.68 / Avg: 557.64 / Max: 672.47 Min: 73.47 / Avg: 590.2 / Max: 753.92 Min: 68.03 / Avg: 617.5 / Max: 820.53
srsRAN Project srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Mbps, More Is Better srsRAN Project 23.10.1-20240325 Test: PDSCH Processor Benchmark, Throughput Total Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P EPYC 9754 2P Xeon 6780E EPYC 9684X EPYC 9654 Xeon 6766E Xeon Platinum 8592+ EPYC 9684X 2P EPYC 9754 Xeon Platinum 8490H EPYC 9654 2P Xeon Max 9468 EPYC 8534PN EPYC 8534P Xeon Platinum 8380 2P Xeon Max 9480 Xeon Platinum 8380 Xeon Max 9480 2P 7K 14K 21K 28K 35K SE +/- 242.64, N = 3 SE +/- 319.61, N = 3 SE +/- 356.81, N = 15 SE +/- 175.78, N = 4 SE +/- 312.17, N = 3 SE +/- 2115.57, N = 15 SE +/- 134.28, N = 4 SE +/- 348.28, N = 4 SE +/- 357.13, N = 12 SE +/- 150.02, N = 4 SE +/- 298.34, N = 15 SE +/- 343.97, N = 3 SE +/- 237.63, N = 3 SE +/- 236.27, N = 4 SE +/- 526.01, N = 15 SE +/- 938.95, N = 15 SE +/- 58.21, N = 3 SE +/- 207.19, N = 5 SE +/- 308.01, N = 15 SE +/- 281.72, N = 15 SE +/- 44.42, N = 3 SE +/- 490.54, N = 15 33010.5 31677.2 30547.6 30236.0 30108.7 29877.1 29804.4 29384.9 27695.2 27281.9 26378.1 25977.4 25194.6 24688.5 22931.8 21286.6 19485.8 19375.9 18943.6 15081.0 13537.7 11866.7 1. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -fno-trapping-math -fno-math-errno -ldl Mbps Per Watt
OpenBenchmarking.org Mbps Per Watt, More Is Better srsRAN Project 23.10.1-20240325 Test: PDSCH Processor Benchmark, Throughput Total EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9654 Xeon Platinum 8592+ EPYC 9754 EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8380 EPYC 9754 2P Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Platinum 8490H 2P Xeon Max 9468 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8380 2P Xeon Max 9480 2P 40 80 120 160 200 165.23 141.87 133.01 121.16 100.26 98.83 96.52 91.42 87.21 75.68 74.22 64.73 58.87 58.04 56.88 52.16 51.12 49.98 41.07 40.98 40.31 19.21 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better srsRAN Project 23.10.1-20240325 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 Xeon Platinum 8592+ EPYC 9654 Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9480 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 130 260 390 520 650 Min: 16.71 / Avg: 117.93 / Max: 142.26 Min: 17.85 / Avg: 136.58 / Max: 166.81 Min: 43.55 / Avg: 205.11 / Max: 263.97 Min: 51.25 / Avg: 229.96 / Max: 288.72 Min: 43.52 / Avg: 245.99 / Max: 312.37 Min: 21.57 / Avg: 261.04 / Max: 306.84 Min: 25.9 / Avg: 266.9 / Max: 379.06 Min: 28.12 / Avg: 276.24 / Max: 341.73 Min: 74.5 / Avg: 281.29 / Max: 379.76 Min: 60.24 / Avg: 283.1 / Max: 369.61 Min: 74.87 / Avg: 289.12 / Max: 373.87 Min: 32.26 / Avg: 321.41 / Max: 408.75 Min: 103.72 / Avg: 426.81 / Max: 506.19 Min: 100.98 / Avg: 470 / Max: 560.27 Min: 104.56 / Avg: 509.97 / Max: 630.46 Min: 41.23 / Avg: 514.74 / Max: 612.62 Min: 68.27 / Avg: 537.03 / Max: 710 Min: 55.25 / Avg: 559.64 / Max: 633.66 Min: 148.45 / Avg: 589.02 / Max: 721.35 Min: 195.95 / Avg: 604.92 / Max: 745.23 Min: 189.97 / Avg: 617.8 / Max: 721.97 Min: 65.52 / Avg: 632.5 / Max: 723.62
Blender Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.1 Blend File: Classroom - Compute: CPU-Only EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 Xeon Platinum 8592+ 2P EPYC 9684X Xeon Platinum 8490H 2P EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E Xeon Platinum 8380 2P Xeon Platinum 8592+ EPYC 8534P Xeon Platinum 8490H EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 30 60 90 120 150 SE +/- 0.05, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.05, N = 3 SE +/- 0.09, N = 3 SE +/- 0.04, N = 3 SE +/- 0.22, N = 3 SE +/- 0.14, N = 3 SE +/- 0.05, N = 3 SE +/- 0.11, N = 3 SE +/- 0.25, N = 3 SE +/- 0.38, N = 3 SE +/- 0.67, N = 3 SE +/- 0.05, N = 3 SE +/- 0.16, N = 3 SE +/- 0.11, N = 3 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.18, N = 3 SE +/- 0.40, N = 3 SE +/- 0.20, N = 3 SE +/- 0.38, N = 3 16.82 18.67 21.12 24.61 27.41 31.65 33.61 35.66 35.91 40.26 40.38 44.71 48.07 52.47 62.54 63.61 67.67 68.61 75.26 76.87 86.67 124.63 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Blender 4.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.21 / Avg: 130.93 / Max: 139.49 Min: 16.44 / Avg: 159.04 / Max: 169.83 Min: 43.66 / Avg: 214.19 / Max: 233.43 Min: 52.8 / Avg: 262.14 / Max: 290.33 Min: 43.65 / Avg: 273.8 / Max: 305.54 Min: 21.92 / Avg: 295.34 / Max: 341.6 Min: 31.02 / Avg: 300.94 / Max: 333.76 Min: 24.56 / Avg: 326.8 / Max: 368.75 Min: 60.22 / Avg: 330.03 / Max: 377.07 Min: 77.25 / Avg: 334.43 / Max: 376.17 Min: 77.77 / Avg: 335.02 / Max: 376.96 Min: 32.63 / Avg: 365.01 / Max: 410.63 Min: 105.51 / Avg: 410.34 / Max: 487.07 Min: 108.2 / Avg: 509.44 / Max: 619.2 Min: 101.27 / Avg: 509.54 / Max: 580.93 Min: 41.08 / Avg: 534.08 / Max: 683.35 Min: 59.51 / Avg: 570.92 / Max: 684.42 Min: 80.44 / Avg: 621.31 / Max: 724.97 Min: 168.91 / Avg: 636.73 / Max: 754.11 Min: 210.08 / Avg: 648.81 / Max: 752.1 Min: 203.45 / Avg: 652.82 / Max: 753.03 Min: 62.19 / Avg: 671.43 / Max: 821.27
Aircrack-ng Aircrack-ng is a tool for assessing WiFi/WLAN network security. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org k/s, More Is Better Aircrack-ng 1.7 EPYC 9684X EPYC 9654 EPYC 9754 Xeon Platinum 8380 2P EPYC 8534P Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon 6780E Xeon 6766E Xeon Platinum 8380 Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P 40K 80K 120K 160K 200K SE +/- 1622.56, N = 3 SE +/- 207.11, N = 3 SE +/- 714.86, N = 3 SE +/- 1234.47, N = 15 SE +/- 338.71, N = 3 SE +/- 840.59, N = 3 SE +/- 1953.08, N = 3 SE +/- 319.60, N = 3 SE +/- 122.31, N = 3 SE +/- 36.42, N = 3 SE +/- 888.98, N = 3 SE +/- 1379.58, N = 4 SE +/- 728.55, N = 3 SE +/- 1027.07, N = 15 SE +/- 1131.21, N = 15 SE +/- 773.70, N = 3 SE +/- 1197.41, N = 15 206619.29 186675.43 177989.89 159709.75 153818.03 153584.20 147469.33 141831.78 134048.96 117588.99 116716.40 114715.79 101665.47 83665.76 65663.82 64181.29 53914.93 1. (CXX) g++ options: -std=gnu++17 -O3 -fvisibility=hidden -fcommon -rdynamic -lnl-3 -lnl-genl-3 -lpcre -lsqlite3 -lpthread -lz -lssl -lcrypto -lhwloc -ldl -lm -pthread k/s Per Watt
OpenBenchmarking.org k/s Per Watt, More Is Better Aircrack-ng 1.7 EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6766E EPYC 9654 Xeon 6780E EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8380 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P 300 600 900 1200 1500 1173.57 1057.15 763.99 725.96 702.97 609.64 585.84 507.73 479.36 433.27 424.59 377.96 333.67 142.86 128.68 112.13 97.52 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Aircrack-ng 1.7 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 120 240 360 480 600 Min: 15.46 / Avg: 120.86 / Max: 145.49 Min: 17.23 / Avg: 145.5 / Max: 168.72 Min: 43.62 / Avg: 158.02 / Max: 182.75 Min: 43.53 / Avg: 191.45 / Max: 221.34 Min: 10.09 / Avg: 232.97 / Max: 273.01 Min: 53.68 / Avg: 239.45 / Max: 291.67 Min: 29.99 / Avg: 265.55 / Max: 307.89 Min: 25.19 / Avg: 302.49 / Max: 376.59 Min: 55.34 / Avg: 307.64 / Max: 375.71 Min: 77.17 / Avg: 309.39 / Max: 375.45 Min: 73.18 / Avg: 311.12 / Max: 377.22 Min: 34.08 / Avg: 352.69 / Max: 410.53 Min: 97.38 / Avg: 478.65 / Max: 580.33 Min: 78.32 / Avg: 498.75 / Max: 586.29 Min: 160.46 / Avg: 552.88 / Max: 630.35 Min: 193.67 / Avg: 585.6 / Max: 670.08 Min: 195.2 / Avg: 585.65 / Max: 659.01
EPYC 9684X 2P: The test run did not produce a result.
EPYC 9654 2P: The test run did not produce a result.
EPYC 9754 2P: The test run did not produce a result.
Xeon 6766E 2P: The test run did not produce a result.
Xeon 6780E 2P: The test run did not produce a result.
WebP Image Encode This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org MP/s, More Is Better WebP Image Encode 1.2.4 Encode Settings: Quality 100, Lossless, Highest Compression Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9654 EPYC 9684X EPYC 9684X 2P Xeon Max 9480 Xeon Max 9468 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Platinum 8380 2P EPYC 8534PN EPYC 8534P EPYC 9754 2P EPYC 9754 Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 0.135 0.27 0.405 0.54 0.675 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 0.60 0.59 0.57 0.57 0.57 0.56 0.55 0.55 0.54 0.54 0.52 0.51 0.50 0.49 0.48 0.48 0.48 0.48 0.39 0.38 0.35 0.35 1. (CC) gcc options: -fvisibility=hidden -O2 -lm MP/s Per Watt
OpenBenchmarking.org MP/s Per Watt, More Is Better WebP Image Encode 1.2.4 Encode Settings: Quality 100, Lossless, Highest Compression EPYC 8534PN EPYC 8534P Xeon 6780E Xeon 6766E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P Xeon Max 9480 Xeon Max 9468 EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.002 0.004 0.006 0.008 0.01 0.009 0.009 0.008 0.007 0.006 0.006 0.005 0.005 0.005 0.004 0.003 0.003 0.003 0.003 0.003 0.003 0.003 0.002 0.002 0.002 0.001 0.001 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better WebP Image Encode 1.2.4 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 EPYC 9754 2P Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 80 160 240 320 400 Min: 43.65 / Avg: 50.5 / Max: 67.21 Min: 43.35 / Avg: 51.07 / Max: 67.23 Min: 15.76 / Avg: 52.31 / Max: 56.63 Min: 15.79 / Avg: 53.42 / Max: 58.48 Min: 20.24 / Avg: 77.06 / Max: 83.97 Min: 27.22 / Avg: 96.75 / Max: 105.57 Min: 50.06 / Avg: 101.4 / Max: 108.28 Min: 30.35 / Avg: 109.81 / Max: 119.47 Min: 104.05 / Avg: 111.28 / Max: 129.11 Min: 105.36 / Avg: 121.62 / Max: 143.58 Min: 34.5 / Avg: 128.71 / Max: 144.36 Min: 60.77 / Avg: 148.44 / Max: 161.11 Min: 80.95 / Avg: 184.91 / Max: 202.94 Min: 39.58 / Avg: 186.91 / Max: 199.67 Min: 83.4 / Avg: 190.72 / Max: 207.25 Min: 97.01 / Avg: 192.53 / Max: 212.04 Min: 57.93 / Avg: 225.88 / Max: 245.29 Min: 61.65 / Avg: 249.88 / Max: 268.66 Min: 79.69 / Avg: 274 / Max: 299.87 Min: 150.62 / Avg: 324.76 / Max: 344.8 Min: 205.98 / Avg: 397.44 / Max: 422.1 Min: 205.48 / Avg: 407.7 / Max: 432.16
libavif avifenc This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 1.0 Encoder Speed: 2 Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 Xeon Platinum 8592+ EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Max 9480 2P Xeon Max 9468 2P Xeon Max 9480 EPYC 9754 Xeon Max 9468 EPYC 8534P EPYC 9754 2P EPYC 8534PN Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 20 40 60 80 100 SE +/- 0.09, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 SE +/- 0.03, N = 3 SE +/- 0.43, N = 3 SE +/- 0.16, N = 3 SE +/- 0.04, N = 3 SE +/- 0.06, N = 3 SE +/- 0.22, N = 3 SE +/- 0.31, N = 3 SE +/- 0.06, N = 3 SE +/- 0.21, N = 3 SE +/- 0.03, N = 3 SE +/- 0.13, N = 3 SE +/- 0.10, N = 3 SE +/- 0.25, N = 3 SE +/- 0.04, N = 3 SE +/- 0.41, N = 3 SE +/- 0.09, N = 3 SE +/- 0.01, N = 3 SE +/- 0.07, N = 3 33.65 34.04 34.15 34.79 35.32 35.40 37.31 37.75 38.14 38.65 39.88 40.35 41.13 41.66 41.92 42.61 47.51 52.23 69.34 71.97 76.47 79.20 1. (CXX) g++ options: -O3 -fPIC -lm CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better libavif avifenc 1.0 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 43.58 / Avg: 59.56 / Max: 150.24 Min: 43.42 / Avg: 65.19 / Max: 183.13 Min: 14.11 / Avg: 69.45 / Max: 130.44 Min: 16.88 / Avg: 75.22 / Max: 155.11 Min: 20.27 / Avg: 107.23 / Max: 247.8 Min: 103.94 / Avg: 125.3 / Max: 229.12 Min: 105.72 / Avg: 131.77 / Max: 264.67 Min: 28.7 / Avg: 137.31 / Max: 316.42 Min: 50.37 / Avg: 150.68 / Max: 290.84 Min: 32.31 / Avg: 157.68 / Max: 383.9 Min: 34.19 / Avg: 161.32 / Max: 375.73 Min: 58.97 / Avg: 196.29 / Max: 374.92 Min: 39.83 / Avg: 217.6 / Max: 369.6 Min: 71.37 / Avg: 235.53 / Max: 369.9 Min: 80.19 / Avg: 235.72 / Max: 375.98 Min: 97.78 / Avg: 258.88 / Max: 544.31 Min: 57.84 / Avg: 267.46 / Max: 494.12 Min: 63.92 / Avg: 296.47 / Max: 540.97 Min: 77.42 / Avg: 350.42 / Max: 724.52 Min: 149.52 / Avg: 390.93 / Max: 714.36 Min: 195.24 / Avg: 458.52 / Max: 733.31 Min: 207.34 / Avg: 461.85 / Max: 707.44
srsRAN Project srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Mbps, More Is Better srsRAN Project 23.10.1-20240325 Test: PUSCH Processor Benchmark, Throughput Total EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P EPYC 9754 EPYC 9684X Xeon Platinum 8592+ 2P EPYC 9654 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 8534P Xeon Platinum 8592+ Xeon 6766E 2P EPYC 8534PN Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6780E Xeon 6766E Xeon Platinum 8380 2K 4K 6K 8K 10K SE +/- 76.45, N = 3 SE +/- 2.29, N = 3 SE +/- 74.95, N = 4 SE +/- 63.64, N = 5 SE +/- 55.17, N = 3 SE +/- 0.12, N = 3 SE +/- 0.20, N = 3 SE +/- 47.68, N = 6 SE +/- 0.15, N = 3 SE +/- 26.47, N = 3 SE +/- 14.40, N = 3 SE +/- 0.17, N = 3 SE +/- 25.25, N = 3 SE +/- 18.09, N = 3 SE +/- 36.60, N = 3 SE +/- 19.02, N = 13 SE +/- 18.43, N = 3 SE +/- 18.30, N = 3 SE +/- 0.09, N = 3 SE +/- 0.00, N = 3 SE +/- 8.17, N = 3 SE +/- 0.07, N = 3 8396.7 6999.7 6835.7 5973.8 5656.9 5493.4 5239.6 4853.6 4386.6 3787.0 3414.2 3288.4 3193.0 3088.7 3067.0 2818.5 2616.9 2514.0 2170.7 2118.6 1953.1 1770.0 1. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -fno-trapping-math -fno-math-errno -ldl Mbps Per Watt
OpenBenchmarking.org Mbps Per Watt, More Is Better srsRAN Project 23.10.1-20240325 Test: PUSCH Processor Benchmark, Throughput Total EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 EPYC 9684X EPYC 9754 2P Xeon 6766E Xeon 6780E Xeon Platinum 8592+ EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H Xeon 6766E 2P Xeon 6780E 2P Xeon Max 9480 Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 2P Xeon Platinum 8380 2P Xeon Max 9468 2P 6 12 18 24 30 26.846 24.996 23.406 20.666 18.052 16.197 15.533 14.180 13.247 12.702 10.976 10.725 10.299 9.956 9.756 8.854 8.808 7.989 7.971 7.552 7.535 6.643 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better srsRAN Project 23.10.1-20240325 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9654 Xeon Platinum 8490H EPYC 9754 Xeon Max 9468 Xeon Max 9480 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 14.51 / Avg: 104.99 / Max: 125.17 Min: 43.71 / Avg: 125.74 / Max: 140.92 Min: 18.17 / Avg: 127.74 / Max: 148.29 Min: 43.57 / Avg: 149.41 / Max: 169.56 Min: 53.15 / Avg: 221.57 / Max: 249.62 Min: 25.89 / Avg: 233.17 / Max: 276.77 Min: 27.31 / Avg: 253.54 / Max: 296.82 Min: 62.91 / Avg: 254.1 / Max: 294.51 Min: 21.5 / Avg: 255.23 / Max: 297.3 Min: 74.32 / Avg: 272.33 / Max: 304.79 Min: 73.76 / Avg: 283.68 / Max: 318.07 Min: 106.26 / Avg: 308.05 / Max: 384.76 Min: 31.36 / Avg: 313.37 / Max: 378.12 Min: 108.83 / Avg: 349.96 / Max: 395.95 Min: 104.84 / Avg: 436.41 / Max: 487.66 Min: 82.2 / Avg: 500.5 / Max: 584.22 Min: 37.54 / Avg: 518.41 / Max: 623.41 Min: 60.3 / Avg: 538.17 / Max: 621.29 Min: 156.44 / Avg: 551.02 / Max: 636.6 Min: 206.4 / Avg: 570.06 / Max: 634.71 Min: 196.65 / Avg: 580.82 / Max: 660.22 Min: 67.45 / Avg: 652.65 / Max: 788.02
Llamafile Mozilla's Llamafile allows distributing and running large language models (LLMs) as a single file. Llamafile aims to make open-source LLMs more accessible to developers and users. Llamafile supports a variety of models, CPUs and GPUs, and other options. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.7 Test: llava-v1.5-7b-q4 - Acceleration: CPU EPYC 9654 EPYC 9684X EPYC 8534P Xeon Platinum 8490H EPYC 8534PN EPYC 9754 Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P EPYC 9754 2P Xeon Platinum 8490H 2P Xeon 6780E Xeon 6766E Xeon Max 9480 Xeon Max 9468 2P Xeon Max 9468 Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon 6780E 2P Xeon 6766E 2P 6 12 18 24 30 SE +/- 0.02, N = 5 SE +/- 0.10, N = 5 SE +/- 0.11, N = 4 SE +/- 0.15, N = 4 SE +/- 0.07, N = 4 SE +/- 0.05, N = 4 SE +/- 0.11, N = 3 SE +/- 0.17, N = 4 SE +/- 0.07, N = 4 SE +/- 0.02, N = 4 SE +/- 0.10, N = 15 SE +/- 0.08, N = 4 SE +/- 0.09, N = 15 SE +/- 0.30, N = 15 SE +/- 0.08, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 SE +/- 0.10, N = 3 SE +/- 1.08, N = 12 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 25.75 25.38 21.00 20.95 20.65 20.64 19.11 16.61 14.47 14.29 12.29 10.94 10.44 9.60 8.56 8.13 7.77 7.67 7.49 7.26 2.99 2.83 Tokens Per Second Per Watt
OpenBenchmarking.org Tokens Per Second Per Watt, More Is Better Llamafile 0.7 Test: llava-v1.5-7b-q4 - Acceleration: CPU EPYC 8534PN EPYC 8534P EPYC 9654 EPYC 9754 EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8490H Xeon Platinum 8592+ Xeon 6766E Xeon 6780E EPYC 9654 2P EPYC 9754 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon 6780E 2P Xeon 6766E 2P 0.0502 0.1004 0.1506 0.2008 0.251 0.223 0.202 0.138 0.126 0.110 0.089 0.080 0.068 0.061 0.056 0.037 0.032 0.031 0.030 0.028 0.027 0.019 0.015 0.014 0.013 0.007 0.007 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Llamafile 0.7 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon 6766E 2P EPYC 9654 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 14.6 / Avg: 92.71 / Max: 131.37 Min: 17.78 / Avg: 104.22 / Max: 157.47 Min: 43.72 / Avg: 139.67 / Max: 256.22 Min: 20.15 / Avg: 163.65 / Max: 315.47 Min: 43.49 / Avg: 172.82 / Max: 332.78 Min: 28.61 / Avg: 186.7 / Max: 318.2 Min: 51.28 / Avg: 215.58 / Max: 284.94 Min: 32.28 / Avg: 231.13 / Max: 402.04 Min: 34.18 / Avg: 244.07 / Max: 354.9 Min: 56.03 / Avg: 262.55 / Max: 376.09 Min: 72.53 / Avg: 283.07 / Max: 370.83 Min: 72.54 / Avg: 288.22 / Max: 376.39 Min: 40.53 / Avg: 343.52 / Max: 619.26 Min: 103.86 / Avg: 385.26 / Max: 509.52 Min: 57.8 / Avg: 388.15 / Max: 621.67 Min: 95.32 / Avg: 409.62 / Max: 565.75 Min: 106.56 / Avg: 429.62 / Max: 646.38 Min: 62.86 / Avg: 462.55 / Max: 790.45 Min: 81.53 / Avg: 512.88 / Max: 707.9 Min: 152.91 / Avg: 537.58 / Max: 745.57 Min: 205.99 / Avg: 559.87 / Max: 716.53 Min: 191.45 / Avg: 566.01 / Max: 704.67
LULESH LULESH is the Livermore Unstructured Lagrangian Explicit Shock Hydrodynamics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org z/s, More Is Better LULESH 2.0.3 Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E 2P Xeon Platinum 8490H 2P Xeon 6766E 2P Xeon Platinum 8592+ EPYC 9684X Xeon 6780E EPYC 9654 Xeon 6766E Xeon Max 9480 EPYC 9754 Xeon Max 9468 Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon Platinum 8380 EPYC 8534P EPYC 8534PN 15K 30K 45K 60K 75K SE +/- 359.29, N = 3 SE +/- 675.87, N = 3 SE +/- 207.54, N = 5 SE +/- 206.20, N = 5 SE +/- 349.39, N = 15 SE +/- 427.75, N = 3 SE +/- 91.45, N = 3 SE +/- 272.44, N = 4 SE +/- 55.60, N = 3 SE +/- 437.51, N = 4 SE +/- 194.65, N = 15 SE +/- 314.85, N = 4 SE +/- 372.12, N = 3 SE +/- 254.25, N = 6 SE +/- 217.67, N = 6 SE +/- 244.17, N = 3 SE +/- 211.12, N = 6 SE +/- 182.38, N = 15 SE +/- 188.07, N = 8 SE +/- 53.94, N = 5 SE +/- 372.63, N = 12 SE +/- 373.78, N = 12 70826.73 54678.49 53269.47 52102.67 48760.82 46126.08 44136.53 43724.06 42946.26 38433.97 27287.26 26864.11 26082.71 26024.60 25831.84 24218.83 23848.54 23072.30 21389.82 17348.43 16949.59 16855.89 1. (CXX) g++ options: -O3 -fopenmp -lm -lmpi_cxx -lmpi z/s Per Watt
OpenBenchmarking.org z/s Per Watt, More Is Better LULESH 2.0.3 Xeon 6766E EPYC 8534PN Xeon Platinum 8592+ Xeon 6780E EPYC 8534P Xeon 6766E 2P Xeon Platinum 8592+ 2P EPYC 9654 Xeon 6780E 2P EPYC 9754 EPYC 9684X EPYC 9654 2P EPYC 9754 2P EPYC 9684X 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 40 80 120 160 200 167.31 162.52 155.84 153.17 149.38 144.08 143.95 143.77 134.87 134.50 130.21 123.49 122.60 122.54 105.60 103.89 100.41 100.20 97.72 88.38 84.83 50.91 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better LULESH 2.0.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon Platinum 8592+ Xeon Max 9480 Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 17.36 / Avg: 103.72 / Max: 135.92 Min: 19.22 / Avg: 113.47 / Max: 157.81 Min: 43.63 / Avg: 155.54 / Max: 229.09 Min: 43.51 / Avg: 175.39 / Max: 270.27 Min: 21.79 / Avg: 180.07 / Max: 311.6 Min: 27.74 / Avg: 181.42 / Max: 285.77 Min: 53.1 / Avg: 204.51 / Max: 281.83 Min: 31.6 / Avg: 209.57 / Max: 345.61 Min: 50.79 / Avg: 218.49 / Max: 329.56 Min: 73.71 / Avg: 237.51 / Max: 333.91 Min: 22.84 / Avg: 246.63 / Max: 365.09 Min: 73.96 / Avg: 248.64 / Max: 365.61 Min: 105.26 / Avg: 298.06 / Max: 439.37 Min: 107.92 / Avg: 327.25 / Max: 513.74 Min: 41.83 / Avg: 376.23 / Max: 647.63 Min: 55.4 / Avg: 394.86 / Max: 598.6 Min: 100.66 / Avg: 420.13 / Max: 555.76 Min: 68.15 / Avg: 446.2 / Max: 721.49 Min: 69.2 / Avg: 492.01 / Max: 729.97 Min: 165.01 / Avg: 494.73 / Max: 717.17 Min: 194.87 / Avg: 531.65 / Max: 728.1 Min: 198.01 / Avg: 533.2 / Max: 722.21
7-Zip Compression This is a test of 7-Zip compression/decompression with its integrated benchmark feature. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 22.01 Test: Decompression Rating EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P EPYC 9754 Xeon 6766E 2P EPYC 9684X EPYC 9654 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon 6780E Xeon 6766E Xeon Max 9468 2P EPYC 8534P Xeon Platinum 8380 2P EPYC 8534PN Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 300K 600K 900K 1200K 1500K SE +/- 14188.09, N = 3 SE +/- 38431.24, N = 3 SE +/- 15794.26, N = 3 SE +/- 903.01, N = 3 SE +/- 291.68, N = 3 SE +/- 2333.15, N = 3 SE +/- 1085.99, N = 3 SE +/- 111.53, N = 3 SE +/- 28320.74, N = 3 SE +/- 185.14, N = 3 SE +/- 6276.88, N = 3 SE +/- 258.19, N = 3 SE +/- 343.82, N = 3 SE +/- 2598.93, N = 3 SE +/- 120.28, N = 3 SE +/- 559.82, N = 3 SE +/- 81.21, N = 3 SE +/- 119.21, N = 3 SE +/- 61.74, N = 3 SE +/- 374.59, N = 3 SE +/- 916.22, N = 3 SE +/- 309.57, N = 3 1340226 1116465 1033277 846335 780871 767313 716065 642626 573410 557054 469131 463768 421063 401707 367989 366534 338355 338146 287862 243347 207738 188345 1. (CXX) g++ options: -lpthread -ldl -O2 -fPIC MIPS Per Watt
OpenBenchmarking.org MIPS Per Watt, More Is Better 7-Zip Compression 22.01 Test: Decompression Rating EPYC 9754 EPYC 9754 2P EPYC 8534PN EPYC 8534P EPYC 9654 EPYC 9684X Xeon 6766E Xeon 6780E EPYC 9654 2P Xeon 6766E 2P EPYC 9684X 2P Xeon 6780E 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Max 9468 2P 700 1400 2100 2800 3500 3140.68 2852.08 2828.90 2768.54 2434.84 2281.23 2265.09 2052.50 2011.14 1954.73 1928.99 1861.55 1119.86 986.76 934.34 908.47 777.16 775.27 773.01 746.10 663.36 637.01 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better 7-Zip Compression 22.01 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 EPYC 9684X Xeon Max 9480 Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 18.26 / Avg: 119.61 / Max: 153.19 Min: 19.72 / Avg: 132.92 / Max: 171.37 Min: 43.69 / Avg: 185.89 / Max: 235.27 Min: 43.58 / Avg: 225.95 / Max: 302.95 Min: 52.56 / Avg: 242.94 / Max: 291.38 Min: 22.44 / Avg: 248.63 / Max: 349.11 Min: 27.34 / Avg: 263.93 / Max: 352.58 Min: 34.7 / Avg: 301.96 / Max: 375.99 Min: 60.14 / Avg: 308.09 / Max: 375.76 Min: 70.41 / Avg: 313.16 / Max: 376.07 Min: 33.61 / Avg: 313.89 / Max: 411.35 Min: 74.78 / Avg: 314.81 / Max: 375.55 Min: 105.13 / Avg: 392.54 / Max: 530.11 Min: 108.05 / Avg: 454.64 / Max: 642.86 Min: 42.06 / Avg: 469.91 / Max: 704.89 Min: 101.47 / Avg: 471.63 / Max: 582.19 Min: 60.89 / Avg: 513.78 / Max: 716.51 Min: 68.73 / Avg: 578.78 / Max: 823.64 Min: 77.81 / Avg: 581.11 / Max: 756.45 Min: 170.4 / Avg: 613.18 / Max: 756.05 Min: 198.7 / Avg: 628.77 / Max: 756.25 Min: 205.27 / Avg: 630.61 / Max: 754.38
OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 22.01 Test: Compression Rating EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon 6780E 2P Xeon Platinum 8490H 2P Xeon 6766E 2P EPYC 9684X EPYC 9754 EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ Xeon 6780E Xeon 6766E Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 EPYC 8534P EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 170K 340K 510K 680K 850K SE +/- 2500.52, N = 3 SE +/- 6869.20, N = 3 SE +/- 3075.95, N = 3 SE +/- 709.87, N = 3 SE +/- 1102.28, N = 3 SE +/- 4200.62, N = 3 SE +/- 7384.79, N = 3 SE +/- 1520.59, N = 3 SE +/- 1188.96, N = 3 SE +/- 719.89, N = 3 SE +/- 3924.87, N = 3 SE +/- 5127.15, N = 3 SE +/- 1982.97, N = 3 SE +/- 933.46, N = 3 SE +/- 610.46, N = 3 SE +/- 1785.84, N = 3 SE +/- 1535.53, N = 3 SE +/- 2258.81, N = 3 SE +/- 268.94, N = 3 SE +/- 206.69, N = 3 SE +/- 1836.55, N = 3 SE +/- 248.73, N = 3 776968 761734 741366 674239 611200 592729 591835 579112 575440 549500 525514 491729 441636 438048 413851 391719 388190 342363 331357 319094 314716 243732 1. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
Algebraic Multi-Grid Benchmark AMG is a parallel algebraic multigrid solver for linear systems arising from problems on unstructured grids. The driver provided with AMG builds linear systems for various 3-dimensional problems. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Figure Of Merit, More Is Better Algebraic Multi-Grid Benchmark 1.2 EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P EPYC 9754 2P EPYC 9684X EPYC 9654 EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 Xeon Platinum 8380 2P 1000M 2000M 3000M 4000M 5000M SE +/- 6340323.98, N = 3 SE +/- 8808919.44, N = 3 SE +/- 9014927.20, N = 3 SE +/- 9164253.67, N = 3 SE +/- 7233468.75, N = 3 SE +/- 2438223.49, N = 3 SE +/- 235140242.43, N = 12 SE +/- 4963892.66, N = 3 SE +/- 1066998.75, N = 3 SE +/- 433397.56, N = 3 SE +/- 4000850.45, N = 3 SE +/- 468935.32, N = 3 SE +/- 1778769.08, N = 3 SE +/- 3359086.30, N = 3 SE +/- 2902028.33, N = 3 SE +/- 291866.98, N = 3 SE +/- 971797.70, N = 3 SE +/- 467864.03, N = 3 4467281333 4324930333 3609646333 3205161000 3156261667 3124402667 2783039083 2376149000 2282802000 2275352667 1814760000 1605097000 1597993333 1588496667 1150052667 1147108000 1073627667 723153533 1. (CC) gcc options: -lparcsr_ls -lparcsr_mv -lseq_mv -lIJ_mv -lkrylov -lHYPRE_utilities -lm -fopenmp -lmpi Figure Of Merit Per Watt
OpenBenchmarking.org Figure Of Merit Per Watt, More Is Better Algebraic Multi-Grid Benchmark 1.2 EPYC 9754 EPYC 8534PN EPYC 9654 EPYC 8534P EPYC 9654 2P EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8592+ EPYC 9754 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 2M 4M 6M 8M 10M 10610537.82 10387087.22 9442032.07 9344903.38 8578578.70 8293777.78 7422966.05 6520382.60 6349445.21 5533330.98 5415151.06 5380716.03 5363427.41 5337195.71 5320019.08 5278683.30 4622831.26 1552824.49 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Algebraic Multi-Grid Benchmark 1.2 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon Platinum 8592+ EPYC 9684X Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 16.12 / Avg: 110.44 / Max: 131.86 Min: 17.02 / Avg: 123.07 / Max: 150.22 Min: 20 / Avg: 214.44 / Max: 294.78 Min: 51.21 / Avg: 232.24 / Max: 283.34 Min: 23.58 / Avg: 241.77 / Max: 304.95 Min: 25.82 / Avg: 285.81 / Max: 370.88 Min: 27.09 / Avg: 286.5 / Max: 375.22 Min: 75.43 / Avg: 295.22 / Max: 373.47 Min: 58.01 / Avg: 296.41 / Max: 368.28 Min: 59.54 / Avg: 297.94 / Max: 371.37 Min: 98.75 / Avg: 465.7 / Max: 552.2 Min: 42.41 / Avg: 502.96 / Max: 635.91 Min: 58.67 / Avg: 504.15 / Max: 640.85 Min: 67.59 / Avg: 553.59 / Max: 733.88 Min: 164.81 / Avg: 591.89 / Max: 729.05 Min: 189.38 / Avg: 593.28 / Max: 741.95 Min: 186.85 / Avg: 600.53 / Max: 731.72 Min: 63.8 / Avg: 601.82 / Max: 781.06
Liquid-DSP LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 1 - Buffer Length: 256 - Filter Length: 512 Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Max 9480 Xeon Max 9480 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Platinum 8380 2P EPYC 9654 EPYC 9654 2P EPYC 9684X EPYC 9684X 2P EPYC 8534P EPYC 9754 2P EPYC 8534PN EPYC 9754 Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 4M 8M 12M 16M 20M SE +/- 9492.69, N = 3 SE +/- 162096.41, N = 15 SE +/- 5238.74, N = 3 SE +/- 17457.89, N = 3 SE +/- 30179.10, N = 3 SE +/- 7571.88, N = 3 SE +/- 8685.88, N = 3 SE +/- 38670.98, N = 3 SE +/- 18478.22, N = 3 SE +/- 11150.49, N = 3 SE +/- 121304.48, N = 3 SE +/- 124115.99, N = 3 SE +/- 134727.30, N = 3 SE +/- 234395.65, N = 15 SE +/- 3464.10, N = 3 SE +/- 11050.39, N = 3 SE +/- 31571.79, N = 3 SE +/- 120841.84, N = 4 SE +/- 88.19, N = 3 SE +/- 133.33, N = 3 SE +/- 115.47, N = 3 SE +/- 33.33, N = 3 18774333 18200533 16906333 16904667 16885333 16884000 16882333 16843333 13608333 13599000 13394667 13390667 13387333 13083133 11290000 11273667 11236333 11174500 6983933 6969967 6284300 6280767 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 1 - Buffer Length: 256 - Filter Length: 512 Xeon Platinum 8592+ EPYC 8534PN EPYC 8534P Xeon Platinum 8490H Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 Xeon Max 9468 Xeon 6766E Xeon Max 9480 EPYC 9684X Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon 6780E 2P EPYC 9754 2P Xeon Max 9468 2P EPYC 9654 2P Xeon Max 9480 2P Xeon 6766E 2P EPYC 9684X 2P 50K 100K 150K 200K 250K 234762.36 232185.74 229953.45 186991.46 158555.72 155096.20 146784.94 145590.23 144779.82 143435.84 136523.06 129024.51 92062.08 75252.11 72744.48 64666.31 63241.27 61488.83 61154.45 60089.56 59249.32 54322.93 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 60 120 180 240 300 Min: 43.55 / Avg: 43.81 / Max: 45.6 Min: 43.37 / Avg: 44.05 / Max: 45.52 Min: 15.14 / Avg: 48.39 / Max: 52.91 Min: 16.02 / Avg: 49.1 / Max: 53.74 Min: 20.45 / Avg: 72.05 / Max: 78.73 Min: 34.71 / Avg: 79.97 / Max: 86.36 Min: 64.97 / Avg: 90.28 / Max: 123.68 Min: 27.87 / Avg: 91.25 / Max: 99.85 Min: 48.5 / Avg: 93.47 / Max: 100.4 Min: 33.62 / Avg: 103.76 / Max: 113.64 Min: 103.81 / Avg: 106.01 / Max: 108.94 Min: 105.16 / Avg: 107.78 / Max: 109.29 Min: 76.32 / Avg: 116.77 / Max: 126.02 Min: 87.22 / Avg: 123.68 / Max: 135.23 Min: 39.35 / Avg: 178.26 / Max: 197.4 Min: 100.48 / Avg: 180.71 / Max: 198.88 Min: 80.57 / Avg: 197.7 / Max: 288.51 Min: 58.4 / Avg: 218.96 / Max: 238.65 Min: 161.66 / Avg: 231.54 / Max: 318.27 Min: 58.81 / Avg: 240.84 / Max: 263.72 Min: 202.94 / Avg: 274.92 / Max: 318.53 Min: 214.89 / Avg: 280.98 / Max: 313.66
NAS Parallel Benchmarks NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.D EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9754 EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8380 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 EPYC 8534P EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 5K 10K 15K 20K 25K SE +/- 602.37, N = 15 SE +/- 1274.85, N = 12 SE +/- 530.31, N = 15 SE +/- 30.19, N = 4 SE +/- 100.23, N = 4 SE +/- 10.93, N = 4 SE +/- 367.88, N = 12 SE +/- 309.03, N = 15 SE +/- 4.83, N = 4 SE +/- 122.47, N = 15 SE +/- 17.53, N = 3 SE +/- 22.33, N = 3 SE +/- 6.77, N = 3 SE +/- 6.53, N = 3 SE +/- 14.68, N = 3 SE +/- 50.01, N = 8 SE +/- 155.10, N = 12 SE +/- 13.82, N = 3 22166.60 22017.75 19414.94 16004.07 14946.76 12793.01 12575.30 11424.05 10970.45 10652.03 8880.82 8089.14 7562.80 6445.09 6416.34 5745.04 5366.69 4479.28 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.D EPYC 9754 2P EPYC 9754 EPYC 8534PN EPYC 8534P EPYC 9654 2P EPYC 9684X 2P EPYC 9654 EPYC 9684X Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Platinum 8380 13 26 39 52 65 58.09 53.74 50.72 49.71 46.56 44.72 43.86 37.12 33.93 29.24 28.72 25.96 23.86 22.77 21.07 20.39 20.17 18.48 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9684X EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 17.1 / Avg: 113.27 / Max: 135.62 Min: 18.8 / Avg: 129.08 / Max: 158.73 Min: 21.19 / Avg: 234.01 / Max: 336.9 Min: 51.37 / Avg: 242.36 / Max: 290.59 Min: 26.47 / Avg: 242.89 / Max: 334.43 Min: 73.35 / Avg: 254.67 / Max: 351.04 Min: 76.21 / Avg: 270.13 / Max: 345.76 Min: 26.35 / Avg: 276.68 / Max: 375.06 Min: 57.58 / Avg: 291.31 / Max: 376.98 Min: 31.93 / Avg: 307.77 / Max: 409.54 Min: 40.5 / Avg: 379.03 / Max: 677.9 Min: 56.44 / Avg: 416.97 / Max: 683.16 Min: 100.12 / Avg: 440.29 / Max: 581.61 Min: 68.69 / Avg: 471.7 / Max: 753.27 Min: 65.8 / Avg: 495.73 / Max: 819.01 Min: 155.56 / Avg: 520.43 / Max: 752.57 Min: 200.97 / Avg: 537.97 / Max: 722.6 Min: 196.31 / Avg: 561.84 / Max: 754.16
ASTC Encoder ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 4.7 Preset: Very Thorough EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon 6780E 2P EPYC 9754 Xeon Max 9480 2P Xeon 6766E 2P EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6780E Xeon Max 9480 Xeon 6766E EPYC 8534P EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8380 5 10 15 20 25 SE +/- 0.1961, N = 15 SE +/- 0.1284, N = 3 SE +/- 0.1965, N = 3 SE +/- 0.0007, N = 3 SE +/- 0.0379, N = 3 SE +/- 0.0314, N = 3 SE +/- 0.0006, N = 3 SE +/- 0.0022, N = 3 SE +/- 0.0386, N = 3 SE +/- 0.0254, N = 3 SE +/- 0.0055, N = 3 SE +/- 0.0006, N = 3 SE +/- 0.0294, N = 3 SE +/- 0.0054, N = 3 SE +/- 0.0005, N = 3 SE +/- 0.0016, N = 3 SE +/- 0.0005, N = 3 SE +/- 0.0007, N = 3 SE +/- 0.0012, N = 3 SE +/- 0.0016, N = 3 SE +/- 0.0026, N = 3 SE +/- 0.0075, N = 3 22.0023 19.4052 17.0098 14.6334 13.7518 12.6220 12.3187 11.8875 11.3573 10.3224 10.2051 9.1844 7.6240 7.0885 6.5166 5.9702 5.8632 5.7420 5.2406 5.1342 4.2935 2.1596 1. (CXX) g++ options: -O3 -flto -pthread MT/s Per Watt
OpenBenchmarking.org MT/s Per Watt, More Is Better ASTC Encoder 4.7 Preset: Very Thorough EPYC 9754 2P EPYC 9754 EPYC 8534PN EPYC 8534P EPYC 9654 2P EPYC 9684X 2P EPYC 9654 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon 6766E Xeon Platinum 8592+ 2P Xeon 6780E Xeon Platinum 8592+ Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Max 9480 2P Xeon Max 9480 Xeon Max 9468 2P Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8380 0.0117 0.0234 0.0351 0.0468 0.0585 0.052 0.047 0.041 0.038 0.036 0.036 0.034 0.033 0.032 0.030 0.030 0.028 0.026 0.026 0.025 0.023 0.020 0.019 0.017 0.016 0.009 0.008 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ASTC Encoder 4.7 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 15.84 / Avg: 128.4 / Max: 151.19 Min: 18.13 / Avg: 151.45 / Max: 177.27 Min: 43.59 / Avg: 194.13 / Max: 235.38 Min: 43.52 / Avg: 249.37 / Max: 311.61 Min: 53.27 / Avg: 255.32 / Max: 291.21 Min: 21.13 / Avg: 262.17 / Max: 365.71 Min: 25.96 / Avg: 270.94 / Max: 337.98 Min: 35.19 / Avg: 294.16 / Max: 378.5 Min: 61.31 / Avg: 301.81 / Max: 376.57 Min: 87.45 / Avg: 313.77 / Max: 375.54 Min: 74.02 / Avg: 315.87 / Max: 377.36 Min: 33 / Avg: 322.42 / Max: 410.63 Min: 104.49 / Avg: 346.65 / Max: 485.81 Min: 39.19 / Avg: 421.17 / Max: 704.37 Min: 106.83 / Avg: 422.58 / Max: 617.03 Min: 58.13 / Avg: 474.96 / Max: 673.2 Min: 97.89 / Avg: 488.29 / Max: 581.33 Min: 71.14 / Avg: 526.63 / Max: 753.88 Min: 64.98 / Avg: 538.44 / Max: 820.03 Min: 154.33 / Avg: 550 / Max: 755.26 Min: 200.66 / Avg: 582.09 / Max: 758.49 Min: 205.14 / Avg: 590.21 / Max: 754.13
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 256 - Model: AlexNet Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Max 9480 2P EPYC 9654 2P Xeon Max 9468 2P EPYC 9754 EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9684X EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon 6780E 2P Xeon Max 9468 Xeon 6766E 2P Xeon Platinum 8380 EPYC 8534P EPYC 8534PN Xeon 6780E Xeon 6766E 400 800 1200 1600 2000 SE +/- 5.06, N = 3 SE +/- 1.86, N = 3 SE +/- 15.82, N = 3 SE +/- 5.81, N = 3 SE +/- 6.01, N = 3 SE +/- 1.15, N = 3 SE +/- 4.01, N = 3 SE +/- 10.23, N = 15 SE +/- 3.03, N = 3 SE +/- 1.54, N = 3 SE +/- 2.58, N = 3 SE +/- 4.27, N = 3 SE +/- 0.54, N = 3 SE +/- 6.82, N = 3 SE +/- 10.51, N = 4 SE +/- 1.21, N = 3 SE +/- 19.06, N = 15 SE +/- 1.28, N = 3 SE +/- 0.99, N = 3 SE +/- 0.41, N = 3 SE +/- 2.12, N = 3 SE +/- 1.03, N = 3 1730.25 1681.44 1560.56 1528.41 1524.74 1479.53 1454.71 1403.46 1340.34 1301.26 1278.58 1277.46 1205.30 1058.99 1010.93 986.04 857.20 852.14 843.63 771.47 668.33 609.67 images/sec Per Watt
OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 256 - Model: AlexNet EPYC 8534PN EPYC 9754 EPYC 8534P EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 EPYC 9654 2P Xeon 6766E Xeon Max 9480 EPYC 9684X 2P Xeon Max 9468 Xeon Platinum 8380 2P Xeon 6780E Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E 2P Xeon 6766E 2P 2 4 6 8 10 7.144 6.976 6.684 5.419 4.808 4.749 4.146 3.829 3.635 3.591 3.535 3.524 3.444 3.248 3.205 3.127 3.126 3.122 2.593 2.515 2.456 2.368 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.16.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 120 240 360 480 600 Min: 18.12 / Avg: 108 / Max: 132.76 Min: 17.24 / Avg: 126.21 / Max: 154.11 Min: 43.66 / Avg: 173.02 / Max: 210.79 Min: 22.22 / Avg: 206.77 / Max: 269.72 Min: 43.57 / Avg: 213.83 / Max: 263.13 Min: 26.89 / Avg: 235.94 / Max: 296.09 Min: 53.46 / Avg: 237.33 / Max: 279.61 Min: 33.64 / Avg: 268.97 / Max: 332.41 Min: 33.53 / Avg: 270.65 / Max: 335.43 Min: 58.81 / Avg: 290.68 / Max: 351.63 Min: 83.56 / Avg: 307.45 / Max: 358.74 Min: 71.72 / Avg: 307.61 / Max: 356.17 Min: 105.44 / Avg: 362.01 / Max: 449.07 Min: 40.95 / Avg: 366.52 / Max: 481.87 Min: 108.29 / Avg: 411.57 / Max: 512.05 Min: 54.66 / Avg: 428.39 / Max: 556.88 Min: 102.12 / Avg: 428.69 / Max: 526.9 Min: 79.54 / Avg: 462.52 / Max: 594.2 Min: 69.6 / Avg: 480.42 / Max: 618.89 Min: 155.31 / Avg: 554.16 / Max: 691.31 Min: 206.23 / Avg: 588.17 / Max: 702.19 Min: 196.2 / Avg: 589.54 / Max: 703.42
miniFE MiniFE Finite Element is an application for unstructured implicit finite element codes. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org CG Mflops, More Is Better miniFE 2.2 Problem Size: Small EPYC 9684X 2P EPYC 9684X EPYC 9654 EPYC 9754 Xeon Platinum 8490H 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9754 2P Xeon 6780E Xeon 6766E Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Platinum 8380 2P EPYC 8534PN EPYC 8534P Xeon Platinum 8380 Xeon Max 9480 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Max 9480 13K 26K 39K 52K 65K SE +/- 2182.93, N = 15 SE +/- 92.04, N = 5 SE +/- 42.89, N = 5 SE +/- 53.07, N = 5 SE +/- 467.78, N = 4 SE +/- 1510.37, N = 15 SE +/- 1633.92, N = 15 SE +/- 2087.13, N = 15 SE +/- 8.42, N = 4 SE +/- 19.87, N = 4 SE +/- 1039.71, N = 15 SE +/- 712.94, N = 12 SE +/- 11.14, N = 4 SE +/- 135.77, N = 4 SE +/- 366.08, N = 4 SE +/- 42.17, N = 4 SE +/- 49.20, N = 4 SE +/- 54.57, N = 4 SE +/- 283.44, N = 3 SE +/- 601.03, N = 12 SE +/- 1097.75, N = 15 SE +/- 572.00, N = 12 59466.6 53167.8 50909.6 50139.9 46877.5 43531.0 42114.6 42113.5 40626.4 40441.0 39580.9 37461.1 35230.1 31830.2 31046.5 24548.8 24526.0 22957.3 21583.2 20091.1 19876.5 18621.2 1. (CXX) g++ options: -O3 -fopenmp -lmpi_cxx -lmpi CG Mflops Per Watt
OpenBenchmarking.org CG Mflops Per Watt, More Is Better miniFE 2.2 Problem Size: Small EPYC 9754 Xeon 6766E Xeon 6780E EPYC 9654 EPYC 9684X EPYC 8534PN EPYC 8534P Xeon 6766E 2P EPYC 9684X 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9654 2P Xeon Platinum 8380 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Max 9468 Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 2P 90 180 270 360 450 392.87 373.28 340.56 334.41 298.90 282.72 254.52 183.43 179.90 165.96 162.67 160.18 156.59 140.84 120.52 114.81 110.48 95.00 79.24 71.52 43.11 39.85 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better miniFE 2.2 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8592+ Xeon 6766E 2P Xeon Platinum 8490H Xeon 6780E 2P Xeon Max 9468 EPYC 9754 2P Xeon Max 9480 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 15.16 / Avg: 86.83 / Max: 132.07 Min: 19.07 / Avg: 96.36 / Max: 147.72 Min: 43.59 / Avg: 108.34 / Max: 211.25 Min: 43.53 / Avg: 119.29 / Max: 232.16 Min: 21.36 / Avg: 127.63 / Max: 261.53 Min: 28.79 / Avg: 152.24 / Max: 300.13 Min: 33.7 / Avg: 177.88 / Max: 364.34 Min: 53.33 / Avg: 190.48 / Max: 282.32 Min: 31.62 / Avg: 198.71 / Max: 331.13 Min: 104.71 / Avg: 215.79 / Max: 487.65 Min: 61.92 / Avg: 224.98 / Max: 377.92 Min: 106.66 / Avg: 225.73 / Max: 522.28 Min: 75.77 / Avg: 253.53 / Max: 365.64 Min: 40.05 / Avg: 258.88 / Max: 518.97 Min: 73.91 / Avg: 260.35 / Max: 366.7 Min: 53.09 / Avg: 309.09 / Max: 579.88 Min: 99.06 / Avg: 326.82 / Max: 539.95 Min: 63.04 / Avg: 330.55 / Max: 718.24 Min: 73.08 / Avg: 366.83 / Max: 700.4 Min: 168.8 / Avg: 424.3 / Max: 732.03 Min: 189.32 / Avg: 498.77 / Max: 730.29 Min: 199.32 / Avg: 500.61 / Max: 723.88
OpenRadioss OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: Cell Phone Drop Test EPYC 9684X EPYC 9684X 2P EPYC 9654 EPYC 9654 2P EPYC 9754 Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9754 2P Xeon Platinum 8592+ Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 EPYC 8534P Xeon Max 9468 EPYC 8534PN Xeon Platinum 8380 10 20 30 40 50 SE +/- 0.19, N = 3 SE +/- 0.05, N = 3 SE +/- 0.26, N = 3 SE +/- 0.05, N = 3 SE +/- 0.26, N = 3 SE +/- 0.32, N = 3 SE +/- 0.01, N = 3 SE +/- 0.05, N = 3 SE +/- 0.05, N = 3 SE +/- 0.18, N = 3 SE +/- 0.10, N = 3 SE +/- 0.07, N = 3 SE +/- 0.08, N = 3 SE +/- 0.02, N = 3 SE +/- 0.18, N = 3 SE +/- 0.08, N = 3 SE +/- 0.35, N = 3 SE +/- 0.21, N = 3 20.45 22.52 22.53 23.53 23.69 26.88 27.51 27.63 27.86 27.90 28.22 31.17 31.49 31.92 33.23 33.75 35.95 43.45 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better OpenRadioss 2023.09.15 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 18.07 / Avg: 113.05 / Max: 140.62 Min: 19.07 / Avg: 127.55 / Max: 161.94 Min: 19.86 / Avg: 210.23 / Max: 333.55 Min: 52.02 / Avg: 232.77 / Max: 290.08 Min: 30.24 / Avg: 238.37 / Max: 343.81 Min: 34.28 / Avg: 272.03 / Max: 374.83 Min: 33.62 / Avg: 276.69 / Max: 409.57 Min: 59.2 / Avg: 289.5 / Max: 374.11 Min: 77.53 / Avg: 297.92 / Max: 375.01 Min: 76.63 / Avg: 302.43 / Max: 377 Min: 41.04 / Avg: 357.25 / Max: 670.96 Min: 97.57 / Avg: 424.07 / Max: 583.23 Min: 58.09 / Avg: 439.02 / Max: 706.97 Min: 60.04 / Avg: 493.1 / Max: 819.11 Min: 71.17 / Avg: 504.55 / Max: 754.24 Min: 172.03 / Avg: 545.45 / Max: 745.98 Min: 199.37 / Avg: 570.84 / Max: 753.68 Min: 206.26 / Avg: 572.24 / Max: 749.99
VVenC VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better VVenC 1.11 Video Input: Bosphorus 4K - Video Preset: Faster EPYC 9684X EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8592+ 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E 4 8 12 16 20 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 SE +/- 0.08, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 SE +/- 0.13, N = 3 SE +/- 0.05, N = 3 SE +/- 0.06, N = 3 SE +/- 0.01, N = 3 SE +/- 0.12, N = 3 SE +/- 0.01, N = 3 SE +/- 0.13, N = 6 SE +/- 0.16, N = 3 SE +/- 0.13, N = 3 SE +/- 0.11, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 17.02 16.73 15.97 15.06 14.79 14.36 14.10 13.97 13.61 13.26 13.10 13.02 12.54 11.97 11.90 11.18 10.39 1. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better VVenC 1.11 Video Input: Bosphorus 4K - Video Preset: Faster EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.0322 0.0644 0.0966 0.1288 0.161 0.143 0.130 0.114 0.108 0.106 0.090 0.080 0.068 0.062 0.060 0.051 0.051 0.038 0.031 0.028 0.023 0.023 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better VVenC 1.11 CPU Power Consumption Monitor Xeon 6766E EPYC 8534PN Xeon 6780E EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 110 220 330 440 550 Min: 43.53 / Avg: 90.8 / Max: 124.66 Min: 17.01 / Avg: 95.43 / Max: 126.76 Min: 43.51 / Avg: 103.62 / Max: 147.39 Min: 18.52 / Avg: 107.71 / Max: 147.66 Min: 20.86 / Avg: 139.36 / Max: 195.74 Min: 29.27 / Avg: 185.29 / Max: 257.81 Min: 54.24 / Avg: 210.37 / Max: 275.5 Min: 34.49 / Avg: 211.98 / Max: 303.77 Min: 34.91 / Avg: 235.39 / Max: 320.85 Min: 60.17 / Avg: 250.97 / Max: 326.74 Min: 74.46 / Avg: 275.77 / Max: 340.06 Min: 73.18 / Avg: 282.13 / Max: 350.27 Min: 100.84 / Avg: 347.13 / Max: 466.54 Min: 77.34 / Avg: 425.45 / Max: 535.49 Min: 149.66 / Avg: 449.44 / Max: 572.09 Min: 204.58 / Avg: 521.13 / Max: 631.04 Min: 209.37 / Avg: 521.79 / Max: 619.75 Video Input: Bosphorus 4K - Video Preset: Faster
EPYC 9684X 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1
EPYC 9654 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1
EPYC 9754 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1
Xeon 6766E 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1
Xeon 6780E 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1
Primesieve Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Primesieve 12.1 Length: 1e13 EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P EPYC 9754 EPYC 9684X EPYC 9654 Xeon Platinum 8592+ 2P Xeon 6780E 2P Xeon Platinum 8490H 2P Xeon 6766E 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 2P EPYC 8534P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6780E EPYC 8534PN Xeon 6766E Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 20 40 60 80 100 SE +/- 0.01, N = 4 SE +/- 0.02, N = 4 SE +/- 0.01, N = 4 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.07, N = 3 SE +/- 0.04, N = 3 SE +/- 0.10, N = 3 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 SE +/- 0.07, N = 3 SE +/- 0.25, N = 3 11.56 12.13 12.63 21.76 23.27 24.24 25.02 26.46 27.11 29.39 30.20 33.92 40.44 45.95 48.41 51.82 52.07 54.39 57.71 58.16 67.32 81.36 1. (CXX) g++ options: -O3 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Primesieve 12.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P EPYC 9754 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 15.99 / Avg: 116.97 / Max: 130.2 Min: 17.27 / Avg: 151.7 / Max: 165.81 Min: 43.71 / Avg: 208.57 / Max: 226.96 Min: 52.11 / Avg: 258.9 / Max: 290.98 Min: 43.4 / Avg: 266.63 / Max: 294.13 Min: 21.63 / Avg: 274.16 / Max: 334.73 Min: 28.63 / Avg: 313.43 / Max: 370.81 Min: 21.47 / Avg: 325.15 / Max: 377.46 Min: 54.67 / Avg: 328.05 / Max: 377.53 Min: 74.48 / Avg: 332.68 / Max: 375.62 Min: 74.26 / Avg: 333.11 / Max: 377.38 Min: 34.15 / Avg: 347.93 / Max: 410.94 Min: 104.03 / Avg: 414.77 / Max: 479.6 Min: 103.1 / Avg: 498.66 / Max: 583 Min: 41.03 / Avg: 500.8 / Max: 682.66 Min: 105.22 / Avg: 507.66 / Max: 599.09 Min: 38.5 / Avg: 568.09 / Max: 750.08 Min: 72.85 / Avg: 610.7 / Max: 755.46 Min: 60.51 / Avg: 618.54 / Max: 821.49 Min: 169.87 / Avg: 623.11 / Max: 754.26 Min: 195.98 / Avg: 641.65 / Max: 754.11 Min: 198.12 / Avg: 644.29 / Max: 754.21
GROMACS The GROMACS (GROningen MAchine for Chemical Simulations) molecular dynamics package testing with the water_GMX50 data. This test profile allows selecting between CPU and GPU-based GROMACS builds. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Ns Per Day, More Is Better GROMACS 2024 Implementation: MPI CPU - Input: water_GMX50_bare EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 5 10 15 20 25 SE +/- 0.224, N = 14 SE +/- 0.067, N = 3 SE +/- 0.189, N = 6 SE +/- 0.046, N = 3 SE +/- 0.019, N = 3 SE +/- 0.052, N = 3 SE +/- 0.032, N = 3 SE +/- 0.025, N = 3 SE +/- 0.032, N = 3 SE +/- 0.034, N = 3 SE +/- 0.009, N = 3 SE +/- 0.036, N = 3 SE +/- 0.010, N = 3 SE +/- 0.007, N = 3 SE +/- 0.019, N = 3 SE +/- 0.004, N = 3 SE +/- 0.004, N = 3 SE +/- 0.018, N = 3 22.525 21.305 20.234 18.540 15.450 15.222 14.127 13.392 13.242 12.514 10.477 8.964 8.762 8.282 7.545 6.499 6.246 4.813 1. (CXX) g++ options: -O3 -lm Ns Per Day Per Watt
OpenBenchmarking.org Ns Per Day Per Watt, More Is Better GROMACS 2024 Implementation: MPI CPU - Input: water_GMX50_bare EPYC 9754 EPYC 8534PN EPYC 9654 EPYC 9684X EPYC 8534P EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8490H 2P Xeon Max 9468 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 0.0119 0.0238 0.0357 0.0476 0.0595 0.053 0.052 0.048 0.048 0.047 0.047 0.040 0.037 0.036 0.035 0.029 0.027 0.026 0.024 0.023 0.022 0.019 0.019 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better GROMACS 2024 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 16.45 / Avg: 120.24 / Max: 142.62 Min: 18.05 / Avg: 137.24 / Max: 164.83 Min: 53.31 / Avg: 248.51 / Max: 290.81 Min: 22.18 / Avg: 250.1 / Max: 348.03 Min: 27.69 / Avg: 260.21 / Max: 348.03 Min: 24.97 / Avg: 289.21 / Max: 376.32 Min: 54.7 / Avg: 306.77 / Max: 376.19 Min: 75.59 / Avg: 311.94 / Max: 375.77 Min: 76.31 / Avg: 314 / Max: 374.89 Min: 33.95 / Avg: 317.95 / Max: 409.49 Min: 40.72 / Avg: 455.51 / Max: 701.31 Min: 104.75 / Avg: 474.08 / Max: 581.34 Min: 56.1 / Avg: 505.98 / Max: 708.92 Min: 75.68 / Avg: 535.3 / Max: 753.28 Min: 166.32 / Avg: 584.52 / Max: 753.32 Min: 65.5 / Avg: 601.18 / Max: 820.89 Min: 201.3 / Avg: 602.79 / Max: 753.69 Min: 205.45 / Avg: 603.46 / Max: 753.72
ASTC Encoder ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 4.7 Preset: Exhaustive EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon 6780E 2P EPYC 9754 Xeon Max 9480 2P Xeon 6766E 2P EPYC 9684X Xeon Max 9468 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6780E Xeon 6766E Xeon Max 9480 EPYC 8534P EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8380 4 8 12 16 20 SE +/- 0.1160, N = 3 SE +/- 0.0170, N = 3 SE +/- 0.0218, N = 3 SE +/- 0.0157, N = 3 SE +/- 0.0164, N = 3 SE +/- 0.0118, N = 3 SE +/- 0.0025, N = 3 SE +/- 0.0016, N = 3 SE +/- 0.0040, N = 3 SE +/- 0.0086, N = 3 SE +/- 0.0033, N = 3 SE +/- 0.0002, N = 3 SE +/- 0.0072, N = 3 SE +/- 0.0024, N = 3 SE +/- 0.0000, N = 3 SE +/- 0.0007, N = 3 SE +/- 0.0011, N = 3 SE +/- 0.0010, N = 3 SE +/- 0.0006, N = 3 SE +/- 0.0003, N = 3 SE +/- 0.0029, N = 3 SE +/- 0.0016, N = 3 14.1038 11.9107 10.6018 8.9091 8.3701 7.9060 7.5573 7.2075 7.1291 6.3027 6.2126 5.6301 4.6303 4.2689 4.0574 3.6505 3.6276 3.5146 3.1993 3.1204 2.6471 1.3319 1. (CXX) g++ options: -O3 -flto -pthread MT/s Per Watt
OpenBenchmarking.org MT/s Per Watt, More Is Better ASTC Encoder 4.7 Preset: Exhaustive EPYC 9754 2P EPYC 9754 EPYC 8534PN EPYC 8534P EPYC 9654 2P Xeon 6766E 2P EPYC 9654 EPYC 9684X 2P EPYC 9684X Xeon 6780E 2P Xeon 6766E Xeon Platinum 8592+ 2P Xeon 6780E Xeon Platinum 8592+ Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Max 9480 2P Xeon Max 9480 Xeon Max 9468 2P Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8380 2P 0.0074 0.0148 0.0222 0.0296 0.037 0.033 0.029 0.025 0.023 0.023 0.021 0.021 0.021 0.020 0.019 0.019 0.017 0.016 0.016 0.015 0.014 0.012 0.012 0.011 0.010 0.005 0.005 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ASTC Encoder 4.7 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 15.2 / Avg: 129.53 / Max: 151.58 Min: 18.31 / Avg: 151.62 / Max: 177.01 Min: 43.7 / Avg: 193.71 / Max: 235.48 Min: 43.55 / Avg: 250.35 / Max: 311.8 Min: 54.67 / Avg: 254.82 / Max: 291.9 Min: 22.1 / Avg: 264.41 / Max: 366.28 Min: 29.24 / Avg: 271.69 / Max: 338.26 Min: 36.03 / Avg: 293.73 / Max: 378.6 Min: 57.84 / Avg: 300.87 / Max: 376.96 Min: 82.06 / Avg: 312.54 / Max: 376.01 Min: 74.57 / Avg: 315.36 / Max: 378.01 Min: 35.44 / Avg: 322.22 / Max: 410.64 Min: 105.14 / Avg: 347.52 / Max: 490.12 Min: 107.83 / Avg: 422.73 / Max: 622.38 Min: 41.13 / Avg: 424.16 / Max: 725.4 Min: 58.38 / Avg: 469.68 / Max: 682.7 Min: 103.02 / Avg: 483.56 / Max: 581.87 Min: 80.68 / Avg: 525.87 / Max: 752.74 Min: 152.83 / Avg: 551.93 / Max: 753.4 Min: 65.5 / Avg: 555.22 / Max: 820.52 Min: 207.67 / Avg: 586.49 / Max: 753.26 Min: 207.97 / Avg: 591.29 / Max: 755.19
uvg266 uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better uvg266 0.4.1 Video Input: Bosphorus 4K - Video Preset: Medium EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P EPYC 9684X EPYC 9754 EPYC 9654 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon 6780E 2P Xeon Max 9468 2P Xeon 6766E 2P EPYC 8534P Xeon Platinum 8490H EPYC 8534PN Xeon 6780E Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Max 9480 Xeon 6766E Xeon Max 9468 Xeon Platinum 8380 10 20 30 40 50 SE +/- 0.46, N = 4 SE +/- 0.23, N = 4 SE +/- 0.28, N = 11 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.01, N = 3 SE +/- 0.05, N = 3 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 SE +/- 0.16, N = 15 SE +/- 0.03, N = 3 SE +/- 0.23, N = 5 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.16, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.00, N = 3 45.26 42.03 38.12 32.74 32.44 31.44 30.37 27.65 24.65 22.59 22.25 22.03 19.25 17.85 17.80 17.29 16.64 16.19 15.68 15.65 15.17 12.56 Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better uvg266 0.4.1 Video Input: Bosphorus 4K - Video Preset: Medium EPYC 8534PN EPYC 9654 EPYC 8534P EPYC 9754 Xeon 6766E Xeon 6780E EPYC 9754 2P EPYC 9684X Xeon 6766E 2P EPYC 9654 2P Xeon 6780E 2P EPYC 9684X 2P Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8380 Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 2P 0.0374 0.0748 0.1122 0.1496 0.187 0.166 0.150 0.148 0.145 0.134 0.124 0.105 0.105 0.094 0.090 0.087 0.086 0.062 0.059 0.057 0.051 0.051 0.051 0.051 0.042 0.039 0.038 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better uvg266 0.4.1 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9654 EPYC 9754 Xeon 6766E 2P Xeon Platinum 8380 Xeon 6780E 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9684X EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 16.52 / Avg: 107.55 / Max: 128.61 Min: 43.61 / Avg: 116.67 / Max: 128.84 Min: 18.2 / Avg: 130.37 / Max: 151.17 Min: 43.45 / Avg: 138.99 / Max: 161.82 Min: 29.07 / Avg: 209.83 / Max: 311.67 Min: 22.31 / Avg: 224.13 / Max: 286.22 Min: 104.75 / Avg: 234.71 / Max: 274.07 Min: 52.84 / Avg: 248.57 / Max: 274.13 Min: 107.22 / Avg: 259.74 / Max: 308.25 Min: 26.08 / Avg: 283.27 / Max: 364.78 Min: 60.4 / Avg: 287.16 / Max: 327.69 Min: 76.55 / Avg: 299.38 / Max: 332.83 Min: 74.3 / Avg: 306.72 / Max: 336.87 Min: 34.79 / Avg: 310.62 / Max: 383.33 Min: 40.54 / Avg: 362.19 / Max: 493.12 Min: 103.64 / Avg: 425.24 / Max: 481.28 Min: 56.62 / Avg: 464.69 / Max: 608.77 Min: 66.05 / Avg: 527.19 / Max: 710.22 Min: 76.25 / Avg: 536.26 / Max: 744.89 Min: 168.73 / Avg: 543.59 / Max: 642.52 Min: 208.68 / Avg: 571.96 / Max: 640.22 Min: 190.07 / Avg: 582.58 / Max: 673.85
John The Ripper This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: WPA PSK EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P EPYC 9754 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9480 2P EPYC 9654 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ Xeon 6780E 2P Xeon Platinum 8490H EPYC 8534P Xeon 6766E 2P EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 Xeon 6780E Xeon 6766E 300K 600K 900K 1200K 1500K SE +/- 4358.90, N = 3 SE +/- 5291.50, N = 3 SE +/- 1333.33, N = 3 SE +/- 478.70, N = 3 SE +/- 15579.94, N = 15 SE +/- 2519.00, N = 3 SE +/- 2683.36, N = 3 SE +/- 788.50, N = 3 SE +/- 921.74, N = 3 SE +/- 5122.51, N = 3 SE +/- 1491.72, N = 3 SE +/- 864.17, N = 3 SE +/- 394.40, N = 3 SE +/- 1599.31, N = 3 SE +/- 321.00, N = 3 SE +/- 187.67, N = 3 SE +/- 79.20, N = 3 SE +/- 1274.00, N = 3 SE +/- 987.33, N = 3 SE +/- 177.03, N = 3 SE +/- 66.16, N = 3 SE +/- 30.00, N = 3 1551000 1294000 1253333 803187 797093 793253 671322 657889 643475 562875 491583 428116 407256 401278 386212 367657 339998 334076 286552 248066 209803 188925 1. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2 Real C/S Per Watt
OpenBenchmarking.org Real C/S Per Watt, More Is Better John The Ripper 2023.03.14 Test: WPA PSK EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9754 2P EPYC 9654 EPYC 9654 2P EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9468 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon 6766E Xeon 6766E 2P Xeon 6780E Xeon 6780E 2P 600 1200 1800 2400 3000 2563.93 2414.27 2304.83 2253.10 2017.72 1941.48 1853.75 1799.50 1374.72 1347.41 1285.88 1261.48 1204.61 1133.97 1100.83 1057.22 1017.12 1004.94 953.29 881.55 815.37 784.52 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better John The Ripper 2023.03.14 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon Max 9468 Xeon 6780E Xeon Max 9480 Xeon Platinum 8490H Xeon Platinum 8592+ EPYC 9654 EPYC 9754 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9654 2P EPYC 9754 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.08 / Avg: 132.61 / Max: 150.07 Min: 18.24 / Avg: 159.97 / Max: 182.75 Min: 43.68 / Avg: 198.18 / Max: 221.69 Min: 53.04 / Avg: 243.89 / Max: 290.93 Min: 66.96 / Avg: 252.7 / Max: 281.63 Min: 43.57 / Avg: 257.31 / Max: 292.39 Min: 76.35 / Avg: 277.33 / Max: 310.37 Min: 56.65 / Avg: 297.81 / Max: 364.7 Min: 23.24 / Avg: 311.42 / Max: 376.46 Min: 29.6 / Avg: 318.91 / Max: 365.41 Min: 22.59 / Avg: 348.48 / Max: 397.44 Min: 33.28 / Avg: 362.14 / Max: 411.02 Min: 108.35 / Avg: 417.06 / Max: 466.57 Min: 99.22 / Avg: 489.17 / Max: 579.96 Min: 109.89 / Avg: 519.11 / Max: 586.15 Min: 186.18 / Avg: 532.41 / Max: 591.91 Min: 193.96 / Avg: 597.63 / Max: 753.44 Min: 64.41 / Avg: 619.88 / Max: 753.08 Min: 165.05 / Avg: 628.83 / Max: 743.61 Min: 59 / Avg: 645.56 / Max: 737.44 Min: 43.51 / Avg: 688.39 / Max: 796.51 Min: 67.44 / Avg: 719.09 / Max: 821.15
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 1 - Model: GoogLeNet Xeon Platinum 8380 EPYC 8534P EPYC 8534PN Xeon Max 9468 EPYC 9684X Xeon Platinum 8490H EPYC 9654 Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8592+ EPYC 9754 Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon 6780E EPYC 9684X 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon 6766E EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 12 24 36 48 60 SE +/- 0.24, N = 7 SE +/- 0.30, N = 15 SE +/- 0.14, N = 6 SE +/- 0.27, N = 7 SE +/- 0.33, N = 15 SE +/- 0.19, N = 6 SE +/- 0.41, N = 15 SE +/- 0.27, N = 6 SE +/- 0.34, N = 15 SE +/- 0.21, N = 6 SE +/- 0.27, N = 15 SE +/- 0.20, N = 15 SE +/- 0.23, N = 12 SE +/- 0.23, N = 15 SE +/- 0.09, N = 4 SE +/- 0.17, N = 15 SE +/- 0.12, N = 6 SE +/- 0.14, N = 15 SE +/- 0.04, N = 4 SE +/- 0.04, N = 3 SE +/- 0.06, N = 9 SE +/- 0.05, N = 7 53.38 36.66 36.48 36.13 34.20 33.56 32.80 31.47 29.18 21.67 21.00 18.24 14.67 13.50 13.14 12.83 12.69 12.42 12.40 8.10 6.72 5.88 images/sec Per Watt
OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 1 - Model: GoogLeNet EPYC 8534PN EPYC 8534P Xeon Platinum 8380 EPYC 9654 EPYC 9684X Xeon Platinum 8490H EPYC 9754 Xeon Max 9468 Xeon 6766E Xeon Platinum 8592+ Xeon 6780E Xeon Max 9480 Xeon Platinum 8380 2P Xeon Max 9468 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon 6780E 2P EPYC 9754 2P Xeon 6766E 2P 0.1458 0.2916 0.4374 0.5832 0.729 0.648 0.607 0.496 0.283 0.261 0.217 0.207 0.204 0.177 0.174 0.172 0.171 0.127 0.043 0.041 0.041 0.038 0.034 0.034 0.032 0.031 0.030 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.16.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 100 200 300 400 500 Min: 15.28 / Avg: 56.3 / Max: 109.74 Min: 14.4 / Avg: 60.4 / Max: 120.48 Min: 43.55 / Avg: 69.97 / Max: 94.67 Min: 43.43 / Avg: 76.44 / Max: 108.57 Min: 20.13 / Avg: 101.49 / Max: 176.75 Min: 50.23 / Avg: 107.65 / Max: 211.46 Min: 27.18 / Avg: 115.93 / Max: 228.86 Min: 26.58 / Avg: 124.36 / Max: 199.6 Min: 30.18 / Avg: 131.02 / Max: 261.02 Min: 57.88 / Avg: 154.91 / Max: 259.01 Min: 74.34 / Avg: 177.15 / Max: 279.76 Min: 71.73 / Avg: 183.7 / Max: 287.38 Min: 104.02 / Avg: 195.59 / Max: 247.83 Min: 105.84 / Avg: 207.14 / Max: 268.89 Min: 98.2 / Avg: 229.87 / Max: 381.04 Min: 39.52 / Avg: 262.36 / Max: 352.55 Min: 57.14 / Avg: 306.45 / Max: 450.83 Min: 73.56 / Avg: 313.09 / Max: 440.43 Min: 61.13 / Avg: 340.38 / Max: 500.24 Min: 148.89 / Avg: 402.29 / Max: 545.86 Min: 186.6 / Avg: 423.67 / Max: 565.69 Min: 191.38 / Avg: 433 / Max: 578.39
Liquid-DSP LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 1 - Buffer Length: 256 - Filter Length: 57 Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9684X EPYC 9654 EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8380 Xeon Platinum 8380 2P EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 8534PN Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 16M 32M 48M 64M 80M SE +/- 599433.09, N = 9 SE +/- 68119.99, N = 3 SE +/- 4910.31, N = 3 SE +/- 5364.49, N = 3 SE +/- 666.67, N = 3 SE +/- 4163.33, N = 3 SE +/- 3844.19, N = 3 SE +/- 822374.71, N = 4 SE +/- 7423.69, N = 3 SE +/- 12719.19, N = 3 SE +/- 34724.31, N = 3 SE +/- 486361.09, N = 3 SE +/- 15982.63, N = 3 SE +/- 178712.37, N = 3 SE +/- 7937.25, N = 3 SE +/- 8089.77, N = 3 SE +/- 4807.40, N = 3 SE +/- 393224.42, N = 3 SE +/- 5206.83, N = 3 SE +/- 11392.00, N = 3 SE +/- 20835.33, N = 3 SE +/- 2185.81, N = 3 73140667 70499000 67130333 67127667 67074667 67064000 66991333 66266000 55455333 55425333 55415667 54954667 54089667 53930333 46270000 46229667 46210667 45486333 37195667 37121333 33444667 33439333 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 1 - Buffer Length: 256 - Filter Length: 57 EPYC 8534PN EPYC 8534P Xeon Platinum 8592+ Xeon 6780E Xeon 6766E Xeon Platinum 8490H EPYC 9754 EPYC 9654 Xeon Max 9468 Xeon Platinum 8380 Xeon Max 9480 EPYC 9684X Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P EPYC 9754 2P EPYC 9654 2P Xeon Max 9468 2P Xeon Max 9480 2P EPYC 9684X 2P 200K 400K 600K 800K 1000K 924272.74 919805.49 859261.41 845147.04 764093.38 741692.03 633663.71 604481.01 575880.01 572291.34 539196.73 529757.56 344751.30 315271.07 310526.31 295293.25 287557.79 256852.19 250729.15 243752.92 235014.31 229312.07 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 60 120 180 240 300 Min: 43.57 / Avg: 43.77 / Max: 45.58 Min: 43.3 / Avg: 44.01 / Max: 45.17 Min: 16 / Avg: 49.21 / Max: 54.17 Min: 17.35 / Avg: 50.24 / Max: 54.91 Min: 21.28 / Avg: 73.02 / Max: 80.63 Min: 33.42 / Avg: 85.12 / Max: 133.8 Min: 62.53 / Avg: 90.43 / Max: 122.85 Min: 26.34 / Avg: 91.69 / Max: 100.53 Min: 50.3 / Avg: 94.51 / Max: 101.21 Min: 31.94 / Avg: 104.68 / Max: 114.13 Min: 103.81 / Avg: 106.07 / Max: 107.92 Min: 104.74 / Avg: 107.68 / Max: 108.89 Min: 75.77 / Avg: 116.57 / Max: 125.17 Min: 84.47 / Avg: 124.5 / Max: 135.18 Min: 40.07 / Avg: 179.99 / Max: 198.16 Min: 97.63 / Avg: 182.63 / Max: 203.03 Min: 56.89 / Avg: 219.18 / Max: 239.46 Min: 79.82 / Avg: 227.03 / Max: 288.68 Min: 162.53 / Avg: 232.97 / Max: 292.63 Min: 64.49 / Avg: 241.66 / Max: 264.36 Min: 206.29 / Avg: 275.13 / Max: 324.32 Min: 203.12 / Avg: 281.97 / Max: 319.46
QuantLib QuantLib is an open-source library/framework around quantitative finance for modeling, trading and risk management scenarios. QuantLib is written in C++ with Boost and its built-in benchmark used reports the QuantLib Benchmark Index benchmark score. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org MFLOPS, More Is Better QuantLib 1.32 Configuration: Single-Threaded Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Max 9468 Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Max 9468 2P EPYC 9684X 2P EPYC 9684X EPYC 9654 EPYC 9654 2P EPYC 9754 EPYC 8534P EPYC 8534PN EPYC 9754 2P Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 800 1600 2400 3200 4000 SE +/- 33.58, N = 3 SE +/- 30.36, N = 12 SE +/- 25.35, N = 3 SE +/- 22.17, N = 3 SE +/- 28.57, N = 3 SE +/- 25.40, N = 3 SE +/- 28.42, N = 3 SE +/- 28.11, N = 3 SE +/- 1.14, N = 3 SE +/- 22.39, N = 3 SE +/- 24.27, N = 3 SE +/- 19.83, N = 3 SE +/- 14.83, N = 3 SE +/- 18.64, N = 3 SE +/- 14.40, N = 3 SE +/- 19.30, N = 3 SE +/- 2.82, N = 3 SE +/- 14.88, N = 3 SE +/- 6.88, N = 3 SE +/- 7.35, N = 3 SE +/- 4.87, N = 3 SE +/- 4.34, N = 3 3600.5 3586.3 3229.1 3225.7 3222.8 3210.1 3209.6 3207.0 3125.4 3114.2 3105.9 3098.5 2596.1 2593.3 2593.0 2590.8 2558.6 2534.5 1960.7 1954.2 1765.6 1760.8 1. (CXX) g++ options: -O3 -march=native -fPIE -pie MFLOPS Per Watt
OpenBenchmarking.org MFLOPS Per Watt, More Is Better QuantLib 1.32 Configuration: Single-Threaded EPYC 8534PN EPYC 8534P Xeon Platinum 8592+ Xeon 6780E Xeon 6766E EPYC 9754 EPYC 9654 Xeon Platinum 8490H EPYC 9684X Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 11 22 33 44 55 50.14 49.12 45.35 43.87 40.30 34.81 33.74 31.94 29.60 27.32 26.68 24.92 18.09 16.48 15.01 14.23 14.07 13.92 13.29 12.80 11.15 11.01 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better QuantLib 1.32 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 Xeon Platinum 8592+ EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8490H EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 70 140 210 280 350 Min: 43.57 / Avg: 43.81 / Max: 45.53 Min: 43.37 / Avg: 44.69 / Max: 46.71 Min: 17.21 / Avg: 51.72 / Max: 57.08 Min: 17.54 / Avg: 52.79 / Max: 57.91 Min: 20.77 / Avg: 74.58 / Max: 81.84 Min: 31.58 / Avg: 79.4 / Max: 103.91 Min: 29.32 / Avg: 92.05 / Max: 101.3 Min: 47.97 / Avg: 92.78 / Max: 99.62 Min: 55.04 / Avg: 100.48 / Max: 155.72 Min: 32.36 / Avg: 105.22 / Max: 115.88 Min: 52.85 / Avg: 106.82 / Max: 109.76 Min: 104.5 / Avg: 108.03 / Max: 111.66 Min: 69.08 / Avg: 120.93 / Max: 183.43 Min: 80.8 / Avg: 129.58 / Max: 191.82 Min: 100.96 / Avg: 183.87 / Max: 210.8 Min: 40.16 / Avg: 184.09 / Max: 202.33 Min: 57.57 / Avg: 217.74 / Max: 241.62 Min: 75.25 / Avg: 238.94 / Max: 292.83 Min: 161.45 / Avg: 241.52 / Max: 334.35 Min: 66.1 / Avg: 244.1 / Max: 269.19 Min: 202.91 / Avg: 287.73 / Max: 404.53 Min: 208.95 / Avg: 292.75 / Max: 399.59
Liquid-DSP LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 64 - Buffer Length: 256 - Filter Length: 512 Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9684X 2P EPYC 9654 2P EPYC 9654 Xeon Platinum 8490H 2P Xeon Platinum 8592+ Xeon Max 9480 2P Xeon Platinum 8490H Xeon Max 9468 2P EPYC 9754 EPYC 9754 2P Xeon Max 9480 Xeon Platinum 8380 2P EPYC 8534P Xeon Max 9468 EPYC 8534PN Xeon Platinum 8380 Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 200M 400M 600M 800M 1000M SE +/- 7562646.32, N = 8 SE +/- 752558.60, N = 3 SE +/- 3039034.93, N = 3 SE +/- 2261241.06, N = 3 SE +/- 1678355.68, N = 3 SE +/- 3717554.45, N = 3 SE +/- 1515552.70, N = 3 SE +/- 6141471.96, N = 3 SE +/- 940265.92, N = 3 SE +/- 4432294.36, N = 3 SE +/- 673003.22, N = 3 SE +/- 958772.36, N = 3 SE +/- 1289603.73, N = 3 SE +/- 749718.47, N = 3 SE +/- 822739.33, N = 3 SE +/- 1611276.24, N = 3 SE +/- 2292977.20, N = 3 SE +/- 147685.85, N = 3 SE +/- 124677.36, N = 3 SE +/- 3610015.39, N = 3 SE +/- 271682.00, N = 3 SE +/- 584646.14, N = 3 883701250 850586667 849340000 824583333 811373333 799496667 783770000 757176667 744410000 732560000 709650000 709393333 689003333 646513333 633860000 632396667 577263333 503253333 398566667 384406667 359576667 349746667 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 64 - Buffer Length: 256 - Filter Length: 512 EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon 6766E EPYC 9684X Xeon Platinum 8592+ Xeon 6780E EPYC 9754 2P Xeon Platinum 8490H Xeon Max 9480 EPYC 9654 2P Xeon Platinum 8380 Xeon Max 9468 EPYC 9684X 2P Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Max 9468 2P 1.2M 2.4M 3.6M 4.8M 6M 5458880.22 5046791.28 3748079.60 3317050.16 2885618.69 2753902.96 2577646.65 2545869.67 2394875.44 2362548.08 2174337.88 2172222.16 2067133.31 1991130.78 1904459.88 1767425.56 1718163.34 1705499.35 1519282.71 1510160.81 1184768.10 1151612.89 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8380 EPYC 9654 EPYC 9754 2P Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 16.94 / Avg: 105.75 / Max: 120.76 Min: 43.66 / Avg: 124.61 / Max: 137.3 Min: 17.47 / Avg: 125.6 / Max: 142.49 Min: 43.52 / Avg: 156.55 / Max: 174.71 Min: 18.89 / Avg: 189.34 / Max: 213.55 Min: 104.54 / Avg: 197.88 / Max: 220.13 Min: 107.13 / Avg: 223.73 / Max: 248.1 Min: 52.9 / Avg: 243.45 / Max: 291.14 Min: 28.29 / Avg: 244.61 / Max: 277.01 Min: 39.32 / Avg: 296.21 / Max: 332.18 Min: 26.5 / Avg: 304.06 / Max: 370.78 Min: 34.73 / Avg: 308.87 / Max: 353.22 Min: 61.24 / Avg: 315.09 / Max: 375.37 Min: 80.39 / Avg: 316.88 / Max: 376.81 Min: 79.61 / Avg: 317.61 / Max: 374.24 Min: 59.09 / Avg: 379.6 / Max: 439.61 Min: 103.26 / Avg: 428.11 / Max: 542.37 Min: 65.79 / Avg: 445.97 / Max: 506.44 Min: 67.01 / Avg: 518.15 / Max: 711.65 Min: 164.45 / Avg: 526.23 / Max: 645.98 Min: 195.75 / Avg: 636.12 / Max: 707.28 Min: 148.91 / Avg: 639.09 / Max: 734.49
SVT-AV1 This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.0 Encoder Mode: Preset 13 - Input: Bosphorus 4K Xeon Platinum 8592+ EPYC 9654 EPYC 9654 2P EPYC 9754 Xeon Platinum 8592+ 2P EPYC 9684X Xeon Platinum 8490H EPYC 9684X 2P EPYC 8534P EPYC 9754 2P Xeon Platinum 8490H 2P EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon 6780E Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 40 80 120 160 200 SE +/- 2.83, N = 15 SE +/- 4.38, N = 15 SE +/- 4.23, N = 15 SE +/- 4.48, N = 15 SE +/- 2.84, N = 15 SE +/- 3.81, N = 15 SE +/- 2.31, N = 15 SE +/- 1.54, N = 15 SE +/- 4.25, N = 15 SE +/- 4.46, N = 15 SE +/- 2.18, N = 15 SE +/- 3.84, N = 15 SE +/- 1.62, N = 15 SE +/- 1.44, N = 15 SE +/- 3.34, N = 15 SE +/- 1.11, N = 15 SE +/- 1.12, N = 7 SE +/- 1.67, N = 15 SE +/- 1.63, N = 15 SE +/- 1.79, N = 15 SE +/- 1.01, N = 8 SE +/- 0.80, N = 15 174.83 172.88 167.22 163.41 162.98 162.48 162.47 161.34 160.79 159.75 154.50 153.42 150.00 144.70 138.72 133.19 124.60 124.26 122.62 116.62 114.54 108.29 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better SVT-AV1 2.0 Encoder Mode: Preset 13 - Input: Bosphorus 4K EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6780E Xeon 6766E EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 EPYC 9754 2P Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 2P Xeon Max 9480 Xeon Max 9468 EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.6239 1.2478 1.8717 2.4956 3.1195 2.773 2.657 1.942 1.691 1.660 1.644 1.366 1.334 1.042 1.033 0.933 0.824 0.821 0.808 0.801 0.773 0.698 0.659 0.632 0.486 0.310 0.308 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better SVT-AV1 2.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8592+ Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9480 Xeon Max 9468 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 100 200 300 400 500 Min: 14.1 / Avg: 55.32 / Max: 107.14 Min: 16.06 / Avg: 60.52 / Max: 126.9 Min: 43.57 / Avg: 69.02 / Max: 130.57 Min: 43.43 / Avg: 73.7 / Max: 142.81 Min: 19.23 / Avg: 84.14 / Max: 169.24 Min: 25.79 / Avg: 105.15 / Max: 208.71 Min: 31.67 / Avg: 118.98 / Max: 227.28 Min: 49.33 / Avg: 128.89 / Max: 223.87 Min: 26.49 / Avg: 131.08 / Max: 275.44 Min: 104.02 / Avg: 131.43 / Max: 193.79 Min: 106.17 / Avg: 142.03 / Max: 219.81 Min: 59.06 / Avg: 155.98 / Max: 273.08 Min: 38.31 / Avg: 171.31 / Max: 283.81 Min: 69.14 / Avg: 187.17 / Max: 307.31 Min: 73.18 / Avg: 187.24 / Max: 299.74 Min: 56.44 / Avg: 206.94 / Max: 347.98 Min: 96.21 / Avg: 219.64 / Max: 381.28 Min: 60.69 / Avg: 231.16 / Max: 380.35 Min: 70.69 / Avg: 247.31 / Max: 459.08 Min: 160.44 / Avg: 317.58 / Max: 460.42 Min: 196.97 / Avg: 398.41 / Max: 555.31 Min: 193.69 / Avg: 401.02 / Max: 565.3
uvg266 uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better uvg266 0.4.1 Video Input: Bosphorus 4K - Video Preset: Slow EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P EPYC 9684X EPYC 9754 EPYC 9654 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9468 2P EPYC 8534P Xeon Platinum 8592+ EPYC 8534PN Xeon Platinum 8490H Xeon 6780E Xeon Max 9480 Xeon Platinum 8380 2P Xeon 6766E Xeon Max 9468 Xeon Platinum 8380 10 20 30 40 50 SE +/- 0.12, N = 4 SE +/- 0.21, N = 4 SE +/- 0.23, N = 4 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.18, N = 3 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 43.39 40.96 37.13 29.18 29.00 28.01 27.11 24.71 22.14 21.63 20.46 19.75 17.04 16.06 15.77 15.71 15.24 13.89 13.79 13.77 13.36 10.93 Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better uvg266 0.4.1 Video Input: Bosphorus 4K - Video Preset: Slow EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6766E Xeon 6780E EPYC 9654 EPYC 9754 2P EPYC 9684X Xeon 6766E 2P EPYC 9654 2P Xeon 6780E 2P EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 2P 0.0331 0.0662 0.0993 0.1324 0.1655 0.147 0.130 0.127 0.117 0.109 0.108 0.099 0.092 0.086 0.086 0.081 0.080 0.055 0.054 0.049 0.045 0.045 0.044 0.044 0.038 0.034 0.033 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better uvg266 0.4.1 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 Xeon 6766E 2P Xeon Platinum 8380 EPYC 9654 Xeon 6780E 2P Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Max 9468 Xeon Max 9480 EPYC 9684X EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 16.27 / Avg: 107.27 / Max: 124.79 Min: 43.62 / Avg: 117.38 / Max: 135.39 Min: 18.85 / Avg: 130.95 / Max: 149.92 Min: 43.49 / Avg: 139.76 / Max: 158.94 Min: 21.98 / Avg: 227.72 / Max: 282.74 Min: 104.27 / Avg: 238.73 / Max: 273.18 Min: 53.14 / Avg: 249.24 / Max: 277.94 Min: 28.73 / Avg: 258.29 / Max: 311.87 Min: 106.34 / Avg: 267.41 / Max: 316.34 Min: 61.29 / Avg: 289.12 / Max: 322.82 Min: 26.18 / Avg: 291.45 / Max: 378.34 Min: 76.03 / Avg: 301.55 / Max: 333.2 Min: 38.16 / Avg: 306.83 / Max: 339.09 Min: 32.45 / Avg: 317.42 / Max: 382.18 Min: 40.4 / Avg: 375.03 / Max: 518.77 Min: 102.97 / Avg: 423.75 / Max: 489.52 Min: 57.99 / Avg: 473.23 / Max: 616.87 Min: 66.68 / Avg: 543.26 / Max: 734.77 Min: 171.75 / Avg: 547.66 / Max: 641.35 Min: 75.78 / Avg: 550.27 / Max: 756.12 Min: 196.37 / Avg: 575.73 / Max: 648.93 Min: 181.76 / Avg: 590.16 / Max: 673.18
Liquid-DSP LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 1 - Buffer Length: 256 - Filter Length: 32 Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Max 9468 Xeon Max 9480 2P Xeon Platinum 8490H Xeon Max 9468 2P Xeon Platinum 8490H 2P EPYC 9654 EPYC 9684X EPYC 9684X 2P EPYC 9654 2P Xeon 6780E Xeon 6780E 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon 6766E Xeon 6766E 2P EPYC 9754 EPYC 8534P EPYC 8534PN EPYC 9754 2P 9M 18M 27M 36M 45M SE +/- 417042.80, N = 5 SE +/- 419600.12, N = 5 SE +/- 3000.00, N = 3 SE +/- 1666.67, N = 3 SE +/- 1452.97, N = 3 SE +/- 1527.53, N = 3 SE +/- 1201.85, N = 3 SE +/- 1154.70, N = 3 SE +/- 7753.14, N = 3 SE +/- 7023.77, N = 3 SE +/- 19139.84, N = 3 SE +/- 13503.09, N = 3 SE +/- 1527.53, N = 3 SE +/- 881.92, N = 3 SE +/- 1000.00, N = 3 SE +/- 577.35, N = 3 SE +/- 0.00, N = 3 SE +/- 881.92, N = 3 SE +/- 6984.11, N = 3 SE +/- 2333.33, N = 3 SE +/- 881.92, N = 3 SE +/- 2185.81, N = 3 40244000 38950600 36539000 36537667 36521333 36506000 36497667 36461000 36068333 36066000 36040000 36033000 35259000 35190667 31995000 31992000 31726000 31705667 29907333 29899667 29899333 29893333 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 1 - Buffer Length: 256 - Filter Length: 32 Xeon 6780E Xeon 6766E EPYC 8534PN EPYC 8534P Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8380 Xeon 6780E 2P Xeon Max 9468 Xeon 6766E 2P Xeon Max 9480 Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 200K 400K 600K 800K 1000K 803665.72 725751.19 596676.31 588721.27 509090.21 411972.01 410323.62 393460.52 344631.62 340868.75 326342.24 318611.94 300148.32 300054.21 215045.22 175115.57 166453.31 165061.92 158420.27 148995.98 133630.14 130400.41 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 60 120 180 240 300 Min: 43.55 / Avg: 43.71 / Max: 45.4 Min: 43.37 / Avg: 43.87 / Max: 45.94 Min: 16.61 / Avg: 50.11 / Max: 55.4 Min: 17.38 / Avg: 50.79 / Max: 56.16 Min: 20.11 / Avg: 72.89 / Max: 80.63 Min: 33.18 / Avg: 79.05 / Max: 85.63 Min: 61.4 / Avg: 88.61 / Max: 120.23 Min: 29.58 / Avg: 91.67 / Max: 100.42 Min: 51.12 / Avg: 93.86 / Max: 101.44 Min: 33.02 / Avg: 104.65 / Max: 115.43 Min: 103.77 / Avg: 105.63 / Max: 107.48 Min: 104.44 / Avg: 107.83 / Max: 110.44 Min: 72.3 / Avg: 114.68 / Max: 123.65 Min: 84.78 / Avg: 121.77 / Max: 132.95 Min: 37.62 / Avg: 179.59 / Max: 198.37 Min: 81.91 / Avg: 181.13 / Max: 198.07 Min: 96.06 / Avg: 182.69 / Max: 203.89 Min: 55.79 / Avg: 218.3 / Max: 239.44 Min: 163.87 / Avg: 230.15 / Max: 297.76 Min: 62.85 / Avg: 241.89 / Max: 266.07 Min: 199.94 / Avg: 273.12 / Max: 314.88 Min: 204.92 / Avg: 280.07 / Max: 323.05
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 64 - Model: GoogLeNet Xeon Platinum 8490H EPYC 9654 Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H 2P EPYC 9684X Xeon Platinum 8592+ EPYC 9754 Xeon Max 9480 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 EPYC 8534P EPYC 8534PN EPYC 9684X 2P Xeon 6780E EPYC 9654 2P Xeon 6766E Xeon 6780E 2P EPYC 9754 2P Xeon 6766E 2P 80 160 240 320 400 SE +/- 1.62, N = 3 SE +/- 4.22, N = 3 SE +/- 1.42, N = 3 SE +/- 3.52, N = 5 SE +/- 0.83, N = 3 SE +/- 1.94, N = 3 SE +/- 2.37, N = 3 SE +/- 1.66, N = 3 SE +/- 1.12, N = 3 SE +/- 3.84, N = 4 SE +/- 2.38, N = 3 SE +/- 0.21, N = 3 SE +/- 0.49, N = 3 SE +/- 0.30, N = 3 SE +/- 0.44, N = 3 SE +/- 2.47, N = 3 SE +/- 0.64, N = 3 SE +/- 2.41, N = 3 SE +/- 0.29, N = 3 SE +/- 0.21, N = 3 SE +/- 1.18, N = 3 SE +/- 2.29, N = 3 380.14 354.59 347.72 341.51 335.58 333.73 331.76 330.26 319.76 314.67 308.95 291.37 282.50 276.93 256.79 248.03 232.51 227.22 214.14 190.80 176.00 168.10 images/sec Per Watt
OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 64 - Model: GoogLeNet EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6766E EPYC 9684X Xeon Platinum 8380 Xeon 6780E Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon 6780E 2P Xeon Max 9480 2P Xeon 6766E 2P EPYC 9654 2P EPYC 9684X 2P EPYC 9754 2P 0.5164 1.0328 1.5492 2.0656 2.582 2.295 2.169 1.603 1.568 1.411 1.378 1.360 1.274 1.261 1.240 1.153 1.142 0.853 0.683 0.633 0.606 0.565 0.553 0.550 0.513 0.503 0.502 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.16.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 120 240 360 480 600 Min: 17.15 / Avg: 111.89 / Max: 138.4 Min: 19.2 / Avg: 127.66 / Max: 158.33 Min: 43.67 / Avg: 157.46 / Max: 196.46 Min: 43.56 / Avg: 187.52 / Max: 236.56 Min: 21.15 / Avg: 196.33 / Max: 257.11 Min: 52.71 / Avg: 224.1 / Max: 274.4 Min: 26.86 / Avg: 226.08 / Max: 291.32 Min: 33.87 / Avg: 226.63 / Max: 286.57 Min: 33.26 / Avg: 259.15 / Max: 327.64 Min: 60.91 / Avg: 275.93 / Max: 349.1 Min: 76.86 / Avg: 291.07 / Max: 351.95 Min: 77.16 / Avg: 292.25 / Max: 352.1 Min: 104.39 / Avg: 305.84 / Max: 369.61 Min: 105.79 / Avg: 337.98 / Max: 420.56 Min: 41.18 / Avg: 350.5 / Max: 413.71 Min: 98.08 / Avg: 407.6 / Max: 516.03 Min: 83.13 / Avg: 426.33 / Max: 534.67 Min: 60.41 / Avg: 443.22 / Max: 530.84 Min: 65.57 / Avg: 493.46 / Max: 594.88 Min: 157.05 / Avg: 524.23 / Max: 643.88 Min: 188.75 / Avg: 558.65 / Max: 667.23 Min: 207.32 / Avg: 563.45 / Max: 671.04
Liquid-DSP LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 64 - Buffer Length: 256 - Filter Length: 57 EPYC 9684X EPYC 9654 EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8592+ EPYC 9754 EPYC 9754 2P Xeon Max 9480 EPYC 8534P Xeon Max 9468 Xeon 6780E Xeon 6780E 2P EPYC 8534PN Xeon Platinum 8380 2P Xeon 6766E Xeon 6766E 2P Xeon Platinum 8380 600M 1200M 1800M 2400M 3000M SE +/- 8476044.68, N = 3 SE +/- 3786085.64, N = 3 SE +/- 8732315.46, N = 3 SE +/- 10590929.03, N = 3 SE +/- 28500058.48, N = 3 SE +/- 32405829.38, N = 3 SE +/- 7076957.92, N = 3 SE +/- 28812627.44, N = 4 SE +/- 27382304.50, N = 5 SE +/- 11426868.92, N = 3 SE +/- 8567444.84, N = 3 SE +/- 17457503.96, N = 3 SE +/- 5493127.02, N = 3 SE +/- 4914717.13, N = 3 SE +/- 2629533.12, N = 3 SE +/- 1361779.88, N = 3 SE +/- 7523814.04, N = 3 SE +/- 7776531.65, N = 3 SE +/- 25737283.81, N = 3 SE +/- 1953913.45, N = 3 SE +/- 4223873.95, N = 3 SE +/- 4885465.97, N = 3 2862000000 2859833333 2772800000 2730233333 2655500000 2620066667 2539200000 2494050000 2491340000 2475500000 2444933333 2374766667 2308966667 2186833333 2138766667 2099033333 2036733333 2026333333 1999266667 1906266667 1852366667 1469166667 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 64 - Buffer Length: 256 - Filter Length: 57 EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8490H EPYC 9754 2P Xeon Platinum 8592+ Xeon Max 9480 EPYC 9654 2P Xeon Max 9468 EPYC 9684X 2P Xeon Platinum 8380 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon Max 9480 2P 4M 8M 12M 16M 20M 18884627.20 16817799.69 14237235.02 12897729.09 12777153.28 10813033.93 9250164.53 9063033.04 8820935.85 8050132.03 8048978.36 8044246.29 7252530.07 6982683.34 6748464.67 6245758.05 5987302.63 5119414.53 4994066.82 4681614.53 3917514.64 3870621.21 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8380 EPYC 9654 EPYC 9754 2P Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 16.16 / Avg: 107.3 / Max: 126.15 Min: 17.61 / Avg: 130.03 / Max: 147.65 Min: 43.66 / Avg: 133.68 / Max: 150 Min: 43.55 / Avg: 162.74 / Max: 182.74 Min: 18.99 / Avg: 191.35 / Max: 215.83 Min: 104.16 / Avg: 204.39 / Max: 225.35 Min: 107.02 / Avg: 230.9 / Max: 255.6 Min: 52.59 / Avg: 245.38 / Max: 291.79 Min: 27.79 / Avg: 264.48 / Max: 300.86 Min: 37.21 / Avg: 295.04 / Max: 331.11 Min: 27.41 / Avg: 307.74 / Max: 373.74 Min: 32.57 / Avg: 309.4 / Max: 348.95 Min: 60.17 / Avg: 315.42 / Max: 375.58 Min: 74.72 / Avg: 316.93 / Max: 377.16 Min: 76.69 / Avg: 318.37 / Max: 376.42 Min: 54.92 / Avg: 391 / Max: 447.88 Min: 101.13 / Avg: 427.05 / Max: 542.12 Min: 64 / Avg: 443.95 / Max: 498.34 Min: 67.62 / Avg: 511.79 / Max: 675.09 Min: 154.62 / Avg: 531.73 / Max: 660.07 Min: 203.4 / Avg: 636.64 / Max: 709.85 Min: 206.33 / Avg: 643.65 / Max: 732.46
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 128 - Buffer Length: 256 - Filter Length: 512 Xeon Platinum 8592+ 2P EPYC 9684X 2P EPYC 9754 EPYC 9684X EPYC 9754 2P Xeon Max 9480 2P EPYC 9654 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 EPYC 8534P Xeon Max 9468 Xeon 6780E Xeon 6780E 2P EPYC 8534PN Xeon 6766E Xeon 6766E 2P Xeon Platinum 8380 300M 600M 900M 1200M 1500M SE +/- 4790383.89, N = 3 SE +/- 3901424.24, N = 3 SE +/- 814452.78, N = 3 SE +/- 2412697.89, N = 3 SE +/- 503322.30, N = 3 SE +/- 2123152.79, N = 3 SE +/- 2338327.99, N = 3 SE +/- 1047748.91, N = 3 SE +/- 2313246.88, N = 3 SE +/- 6842108.19, N = 3 SE +/- 11560043.25, N = 5 SE +/- 8061118.62, N = 3 SE +/- 1740596.96, N = 3 SE +/- 1345523.44, N = 3 SE +/- 390953.25, N = 3 SE +/- 681281.47, N = 3 SE +/- 287923.60, N = 3 SE +/- 866544.86, N = 3 SE +/- 431058.13, N = 3 SE +/- 305632.31, N = 3 SE +/- 1072572.19, N = 3 SE +/- 912164.70, N = 3 1493833333 1394733333 1369800000 1368733333 1349800000 1206066667 1199966667 1158166667 1152933333 1126466667 1091460000 981090000 918566667 882200000 809916667 800186667 796340000 777940000 731756667 719626667 707053333 539226667 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 128 - Buffer Length: 256 - Filter Length: 512 EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ EPYC 9754 2P Xeon 6766E Xeon 6780E EPYC 9654 2P Xeon Platinum 8490H EPYC 9684X 2P Xeon Max 9480 Xeon 6766E 2P Xeon Max 9468 Xeon 6780E 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 1.4M 2.8M 4.2M 5.6M 7M 6364836.35 5867538.62 5183635.11 4367851.07 3778092.65 3740373.11 3727692.53 3673478.12 3117578.50 2957086.75 2929507.99 2805690.78 2800439.16 2621658.59 2528910.04 2466244.28 2443783.02 2212405.84 2021740.47 1966021.92 1899608.97 1702250.87 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 EPYC 9654 Xeon 6766E 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon 6780E 2P Xeon Max 9468 EPYC 9754 2P EPYC 9684X EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 16.09 / Avg: 114.97 / Max: 130.99 Min: 18.57 / Avg: 138.03 / Max: 156.97 Min: 43.64 / Avg: 195.9 / Max: 223.9 Min: 52.09 / Avg: 243.73 / Max: 291.23 Min: 43.62 / Avg: 255.44 / Max: 293.66 Min: 21.47 / Avg: 264.25 / Max: 302.73 Min: 26.08 / Avg: 265.16 / Max: 303.52 Min: 104.69 / Avg: 269.7 / Max: 304.35 Min: 27.09 / Avg: 308.24 / Max: 376.57 Min: 60 / Avg: 313.56 / Max: 376 Min: 75.73 / Avg: 315.02 / Max: 376.2 Min: 107.89 / Avg: 315.44 / Max: 358.75 Min: 77.21 / Avg: 316.42 / Max: 377.29 Min: 40.86 / Avg: 362.1 / Max: 421.05 Min: 32.21 / Avg: 362.28 / Max: 409.42 Min: 53.51 / Avg: 405.79 / Max: 476.68 Min: 102.29 / Avg: 485.27 / Max: 581.46 Min: 65.97 / Avg: 497.11 / Max: 585.32 Min: 168.69 / Avg: 572.97 / Max: 696.31 Min: 70.15 / Avg: 611.28 / Max: 739.63 Min: 206.37 / Avg: 634.9 / Max: 748.44 Min: 197.48 / Avg: 641.19 / Max: 750.61
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 256 - Buffer Length: 256 - Filter Length: 512 EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9754 EPYC 9754 2P Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E 2P EPYC 9654 Xeon 6766E 2P Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon 6780E Xeon 6766E Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 500M 1000M 1500M 2000M 2500M SE +/- 14870701.99, N = 3 SE +/- 11753202.87, N = 3 SE +/- 14465476.14, N = 3 SE +/- 1300000.00, N = 3 SE +/- 1650252.51, N = 3 SE +/- 5630373.98, N = 3 SE +/- 2676440.42, N = 3 SE +/- 6728628.72, N = 3 SE +/- 4238841.56, N = 3 SE +/- 6411708.04, N = 3 SE +/- 520683.31, N = 3 SE +/- 2198737.01, N = 3 SE +/- 240370.09, N = 3 SE +/- 11053858.05, N = 3 SE +/- 1722723.94, N = 3 SE +/- 740202.67, N = 3 SE +/- 409267.64, N = 3 SE +/- 320537.74, N = 3 SE +/- 569766.03, N = 3 SE +/- 776881.23, N = 3 SE +/- 170391.71, N = 3 SE +/- 4629652.01, N = 3 2394333333 2129533333 2027000000 1744700000 1734800000 1616966667 1570400000 1518933333 1444066667 1437100000 1362166667 1306733333 1178766667 1061033333 1020833333 944550000 940270000 848376667 838530000 821013333 741740000 542606667 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 256 - Buffer Length: 256 - Filter Length: 512 EPYC 8534PN EPYC 9754 EPYC 8534P EPYC 9654 EPYC 9754 2P EPYC 9684X EPYC 9654 2P Xeon 6766E Xeon Platinum 8592+ EPYC 9684X 2P Xeon 6780E Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon 6766E 2P Xeon Max 9480 Xeon 6780E 2P Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 1.4M 2.8M 4.2M 5.6M 7M 6532499.05 6413920.52 5998206.24 4853033.94 4657018.09 4387111.54 4001237.14 3974344.33 3805850.46 3603186.16 3349035.56 3272343.49 3267678.46 3182607.79 2982249.17 2808259.47 2666847.49 2571656.77 2388645.74 2269719.80 2257184.32 2190947.55 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon 6780E Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9684X EPYC 9754 2P Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P 140 280 420 560 700 Min: 17.93 / Avg: 113.55 / Max: 130.59 Min: 17.86 / Avg: 136.88 / Max: 157.55 Min: 43.63 / Avg: 213.46 / Max: 250.49 Min: 53.55 / Avg: 240.39 / Max: 291.06 Min: 20.69 / Avg: 272.02 / Max: 314.58 Min: 29.69 / Avg: 280.68 / Max: 320.8 Min: 43.73 / Avg: 280.76 / Max: 332.63 Min: 25.3 / Avg: 309.72 / Max: 376.79 Min: 60.01 / Avg: 312.4 / Max: 376.93 Min: 74.44 / Avg: 314.43 / Max: 376.72 Min: 77.34 / Avg: 316.72 / Max: 376.86 Min: 35.59 / Avg: 357.96 / Max: 411.45 Min: 40.06 / Avg: 372.51 / Max: 431.46 Min: 108.01 / Avg: 410.59 / Max: 495.66 Min: 104.84 / Avg: 484.28 / Max: 583.22 Min: 109.82 / Avg: 511.74 / Max: 621.27 Min: 53.43 / Avg: 532.22 / Max: 613.02 Min: 70.53 / Avg: 619.43 / Max: 752.7 Min: 157.59 / Avg: 628.76 / Max: 754.2 Min: 198.99 / Avg: 635.9 / Max: 752.9 Min: 210.59 / Avg: 636.23 / Max: 755.25 Min: 67.21 / Avg: 664.5 / Max: 779.27
John The Ripper This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: bcrypt EPYC 9754 2P Xeon 6780E 2P EPYC 9684X 2P Xeon 6766E 2P EPYC 9654 2P EPYC 9754 Xeon 6780E Xeon Platinum 8592+ 2P EPYC 9684X Xeon 6766E Xeon Platinum 8490H 2P EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ EPYC 8534P Xeon Platinum 8490H EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 90K 180K 270K 360K 450K SE +/- 4958.68, N = 4 SE +/- 2096.39, N = 3 SE +/- 1315.55, N = 3 SE +/- 801.76, N = 3 SE +/- 710.46, N = 3 SE +/- 1.33, N = 3 SE +/- 87.49, N = 3 SE +/- 288.38, N = 3 SE +/- 154.46, N = 3 SE +/- 32.84, N = 3 SE +/- 292.24, N = 3 SE +/- 20.00, N = 3 SE +/- 576.80, N = 3 SE +/- 866.68, N = 3 SE +/- 12.67, N = 3 SE +/- 46.28, N = 3 SE +/- 22.28, N = 3 SE +/- 10.67, N = 3 SE +/- 0.00, N = 3 SE +/- 8.67, N = 3 SE +/- 8.84, N = 3 SE +/- 6.33, N = 3 412723 402275 376268 363133 320527 220188 211311 206778 192898 190058 183603 166541 155959 133866 113219 105781 102222 94551 93472 79144 67875 56939 1. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2 Real C/S Per Watt
OpenBenchmarking.org Real C/S Per Watt, More Is Better John The Ripper 2023.03.14 Test: bcrypt EPYC 8534PN Xeon 6766E Xeon 6766E 2P EPYC 9754 EPYC 8534P Xeon 6780E Xeon 6780E 2P EPYC 9754 2P EPYC 9654 EPYC 9654 2P EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 200 400 600 800 1000 920.41 885.93 816.03 759.60 749.17 746.76 712.79 705.60 606.47 573.89 531.63 523.96 342.68 334.53 305.82 292.69 273.70 256.68 256.35 241.56 233.71 231.51
Liquid-DSP LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 128 - Buffer Length: 256 - Filter Length: 57 EPYC 9684X 2P EPYC 9684X EPYC 9754 EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9654 Xeon 6780E Xeon 6780E 2P Xeon Max 9480 2P Xeon 6766E Xeon Platinum 8490H 2P Xeon 6766E 2P Xeon Max 9468 2P Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 8534P Xeon Max 9480 Xeon Platinum 8380 2P EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 1000M 2000M 3000M 4000M 5000M SE +/- 11892621.43, N = 3 SE +/- 6778479.02, N = 3 SE +/- 18691531.06, N = 3 SE +/- 14606429.33, N = 3 SE +/- 51282832.98, N = 4 SE +/- 4098915.85, N = 3 SE +/- 20120967.50, N = 3 SE +/- 1909915.65, N = 3 SE +/- 13354940.83, N = 3 SE +/- 14772609.79, N = 3 SE +/- 6887912.12, N = 3 SE +/- 13655808.69, N = 3 SE +/- 2709448.49, N = 3 SE +/- 25888414.40, N = 3 SE +/- 5260228.13, N = 3 SE +/- 3116800.35, N = 3 SE +/- 3527510.93, N = 3 SE +/- 3119294.79, N = 3 SE +/- 11686221.71, N = 3 SE +/- 6398003.16, N = 3 SE +/- 2030052.00, N = 3 SE +/- 1549551.91, N = 3 4759366667 4755333333 4714500000 4626566667 4450525000 4280933333 4256300000 4136766667 4068666667 3896500000 3736300000 3723666667 3712133333 3596400000 3503200000 3181266667 2982800000 2972400000 2689033333 2687066667 2670033333 1554666667 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 128 - Buffer Length: 256 - Filter Length: 57 EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8592+ EPYC 9654 2P Xeon Platinum 8490H EPYC 9684X 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 2P 5M 10M 15M 20M 25M 22733164.52 20620501.38 17916149.91 16705633.01 15275772.40 14722432.57 13110978.99 13085490.53 12284401.92 12234041.51 11273487.11 10089896.41 10089438.81 9563226.72 9312055.82 8437038.99 7209189.04 6447021.00 6364402.84 6058163.97 5639264.93 5516263.53 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E EPYC 9754 Xeon 6766E 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6780E 2P EPYC 9684X EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 17.47 / Avg: 118.2 / Max: 136.58 Min: 17.85 / Avg: 144.65 / Max: 162.72 Min: 43.75 / Avg: 208.54 / Max: 241.22 Min: 51.05 / Avg: 244.28 / Max: 290.35 Min: 43.58 / Avg: 270.81 / Max: 310.28 Min: 21.64 / Avg: 282.21 / Max: 320.93 Min: 104.64 / Avg: 283.13 / Max: 314.71 Min: 29.04 / Avg: 289.1 / Max: 326.36 Min: 24.5 / Avg: 310.75 / Max: 376.14 Min: 32.16 / Avg: 315.31 / Max: 376.7 Min: 75.83 / Avg: 316.47 / Max: 376.18 Min: 79.18 / Avg: 319.2 / Max: 376.38 Min: 107.77 / Avg: 331.21 / Max: 370.24 Min: 35.08 / Avg: 363.41 / Max: 409.45 Min: 40.47 / Avg: 378.17 / Max: 438.15 Min: 57.66 / Avg: 424.28 / Max: 495.38 Min: 101.61 / Avg: 487.47 / Max: 581.03 Min: 60.6 / Avg: 497.67 / Max: 581.47 Min: 158.24 / Avg: 577.58 / Max: 700.7 Min: 69.61 / Avg: 617.34 / Max: 730.41 Min: 204.5 / Avg: 637.74 / Max: 753.77 Min: 195.26 / Avg: 643.18 / Max: 753.46
SVT-AV1 This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.0 Encoder Mode: Preset 4 - Input: Bosphorus 4K EPYC 9684X EPYC 9684X 2P EPYC 9654 EPYC 9654 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9754 EPYC 9754 2P Xeon Platinum 8490H EPYC 8534P EPYC 8534PN Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9468 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 2 4 6 8 10 SE +/- 0.021, N = 3 SE +/- 0.040, N = 3 SE +/- 0.019, N = 3 SE +/- 0.046, N = 3 SE +/- 0.005, N = 3 SE +/- 0.044, N = 3 SE +/- 0.024, N = 3 SE +/- 0.004, N = 3 SE +/- 0.030, N = 3 SE +/- 0.007, N = 3 SE +/- 0.015, N = 3 SE +/- 0.049, N = 3 SE +/- 0.047, N = 3 SE +/- 0.027, N = 3 SE +/- 0.042, N = 3 SE +/- 0.015, N = 3 SE +/- 0.014, N = 3 SE +/- 0.014, N = 3 SE +/- 0.003, N = 3 SE +/- 0.041, N = 3 SE +/- 0.002, N = 3 SE +/- 0.023, N = 3 8.595 8.513 8.422 8.349 7.876 7.360 7.283 7.153 7.061 6.939 6.727 6.698 6.572 6.490 6.052 5.991 5.438 5.437 4.350 4.040 3.972 3.710 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better SVT-AV1 2.0 Encoder Mode: Preset 4 - Input: Bosphorus 4K EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon 6766E Xeon 6780E EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 EPYC 9684X 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.0205 0.041 0.0615 0.082 0.1025 0.091 0.084 0.067 0.060 0.056 0.055 0.054 0.040 0.034 0.033 0.032 0.031 0.029 0.028 0.027 0.027 0.027 0.022 0.021 0.018 0.013 0.013 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better SVT-AV1 2.0 CPU Power Consumption Monitor Xeon 6766E EPYC 8534PN Xeon 6780E EPYC 8534P EPYC 9754 Xeon 6766E 2P EPYC 9654 Xeon 6780E 2P EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 110 220 330 440 550 Min: 43.55 / Avg: 71.01 / Max: 131.44 Min: 15.18 / Avg: 74.03 / Max: 115.16 Min: 43.43 / Avg: 78.63 / Max: 157.97 Min: 16.63 / Avg: 82.61 / Max: 136.8 Min: 20.56 / Avg: 108.62 / Max: 213.33 Min: 104.17 / Avg: 135.78 / Max: 199.44 Min: 26.67 / Avg: 140.24 / Max: 261.59 Min: 105.45 / Avg: 146.79 / Max: 227.06 Min: 33.04 / Avg: 159.73 / Max: 314.08 Min: 52.81 / Avg: 173.53 / Max: 276.87 Min: 34.17 / Avg: 196.16 / Max: 355 Min: 59.43 / Avg: 204.79 / Max: 319.21 Min: 40.1 / Avg: 214.15 / Max: 316.2 Min: 73.83 / Avg: 242.37 / Max: 350.93 Min: 73.14 / Avg: 245.81 / Max: 351.66 Min: 97.29 / Avg: 264.19 / Max: 369.25 Min: 57.39 / Avg: 264.94 / Max: 394.93 Min: 64.63 / Avg: 294.39 / Max: 468.18 Min: 78.08 / Avg: 333.26 / Max: 452.51 Min: 168.49 / Avg: 378.27 / Max: 506.21 Min: 209.94 / Avg: 456.54 / Max: 621.56 Min: 198.73 / Avg: 457.13 / Max: 618.95
Embree Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better Embree 4.3 Binary: Pathtracer ISPC - Model: Asian Dragon Obj EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon 6780E 2P EPYC 9754 EPYC 9684X Xeon Platinum 8490H 2P Xeon 6766E 2P EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon 6780E Xeon Platinum 8490H Xeon 6766E Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 50 100 150 200 250 SE +/- 1.72, N = 15 SE +/- 3.26, N = 12 SE +/- 1.58, N = 15 SE +/- 0.29, N = 4 SE +/- 0.09, N = 3 SE +/- 0.07, N = 4 SE +/- 0.11, N = 4 SE +/- 0.22, N = 4 SE +/- 0.22, N = 3 SE +/- 0.02, N = 4 SE +/- 0.89, N = 15 SE +/- 0.76, N = 3 SE +/- 0.35, N = 3 SE +/- 0.10, N = 3 SE +/- 0.09, N = 3 SE +/- 0.09, N = 3 SE +/- 0.03, N = 3 SE +/- 0.09, N = 3 SE +/- 0.08, N = 3 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 210.44 200.16 175.60 173.22 136.22 133.93 133.48 132.28 125.11 113.28 112.72 104.84 97.19 84.80 83.73 79.38 76.53 71.83 64.96 64.33 63.34 47.64 Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Embree 4.3 Binary: Pathtracer ISPC - Model: Asian Dragon Obj EPYC 9754 EPYC 9754 2P Xeon 6766E Xeon 6780E EPYC 9684X EPYC 8534PN EPYC 9654 Xeon 6780E 2P EPYC 8534P Xeon 6766E 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Platinum 8380 2P Xeon Max 9468 Xeon Platinum 8380 Xeon Max 9480 2P Xeon Max 9468 2P 0.2201 0.4402 0.6603 0.8804 1.1005 0.978 0.759 0.746 0.719 0.716 0.713 0.693 0.643 0.642 0.635 0.602 0.579 0.575 0.523 0.368 0.320 0.291 0.272 0.263 0.261 0.240 0.222 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Embree 4.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 17.83 / Avg: 88.8 / Max: 143.53 Min: 17.71 / Avg: 100.16 / Max: 169.93 Min: 43.56 / Avg: 102.53 / Max: 218.97 Min: 43.49 / Avg: 116.43 / Max: 260.73 Min: 19.95 / Avg: 136.96 / Max: 327.05 Min: 28.7 / Avg: 163.54 / Max: 333.81 Min: 54.43 / Avg: 182.65 / Max: 290.73 Min: 26.36 / Avg: 185.86 / Max: 375.76 Min: 32.74 / Avg: 186.36 / Max: 410.63 Min: 104.6 / Avg: 197.13 / Max: 492.74 Min: 106.69 / Avg: 211.77 / Max: 566.07 Min: 61.45 / Avg: 215.69 / Max: 377.02 Min: 86.41 / Avg: 246.7 / Max: 376.83 Min: 73.93 / Avg: 247.01 / Max: 377.15 Min: 36.73 / Avg: 263.84 / Max: 656.51 Min: 66.3 / Avg: 299.17 / Max: 755.1 Min: 56.83 / Avg: 305.41 / Max: 649.39 Min: 101.28 / Avg: 312.13 / Max: 580.75 Min: 65.49 / Avg: 349.82 / Max: 819.75 Min: 164.62 / Avg: 413.97 / Max: 758.06 Min: 187.33 / Avg: 469.55 / Max: 757.04 Min: 193.13 / Avg: 472.12 / Max: 755.04
John The Ripper This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: Blowfish EPYC 9754 2P Xeon 6780E 2P EPYC 9684X 2P Xeon 6766E 2P EPYC 9654 2P EPYC 9754 Xeon 6780E Xeon Platinum 8592+ 2P EPYC 9684X Xeon 6766E Xeon Platinum 8490H 2P EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ EPYC 8534P Xeon Platinum 8490H EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 90K 180K 270K 360K 450K SE +/- 2397.15, N = 3 SE +/- 482.17, N = 3 SE +/- 454.86, N = 3 SE +/- 353.58, N = 3 SE +/- 3455.48, N = 3 SE +/- 1.33, N = 3 SE +/- 14.67, N = 3 SE +/- 649.08, N = 3 SE +/- 98.33, N = 3 SE +/- 174.44, N = 3 SE +/- 348.45, N = 3 SE +/- 38.33, N = 3 SE +/- 169.28, N = 3 SE +/- 116.69, N = 3 SE +/- 14.62, N = 3 SE +/- 79.75, N = 3 SE +/- 13.72, N = 3 SE +/- 11.39, N = 3 SE +/- 20.54, N = 3 SE +/- 8.67, N = 3 SE +/- 17.33, N = 3 SE +/- 6.33, N = 3 415412 404739 374589 363300 316492 220188 211449 205499 192574 190209 183583 166541 156658 134562 113223 105831 102198 94513 93473 79152 67893 56939 1. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2 Real C/S Per Watt
OpenBenchmarking.org Real C/S Per Watt, More Is Better John The Ripper 2023.03.14 Test: Blowfish Xeon 6766E Xeon 6766E 2P EPYC 8534PN EPYC 9754 EPYC 8534P Xeon 6780E Xeon 6780E 2P EPYC 9754 2P EPYC 9654 EPYC 9654 2P EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 200 400 600 800 1000 885.52 818.19 805.43 757.83 753.26 747.26 720.56 711.80 605.49 568.71 531.08 527.64 339.84 332.35 305.95 291.56 274.74 258.82 257.14 243.04 233.58 231.64 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better John The Ripper 2023.03.14 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon Max 9468 EPYC 9654 Xeon 6780E Xeon Max 9480 EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9684X Xeon 6766E 2P Xeon 6766E 2P Full Xeon Platinum 8380 2P Xeon Max 9468 2P EPYC 9654 2P Xeon 6780E 2P EPYC 9754 2P Xeon Max 9480 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P 140 280 420 560 700 Min: 15.84 / Avg: 101.56 / Max: 134.16 Min: 18.33 / Avg: 135.67 / Max: 154.48 Min: 43.68 / Avg: 214.53 / Max: 240.46 Min: 52.63 / Avg: 243.63 / Max: 290.65 Min: 70.16 / Avg: 264.03 / Max: 291.4 Min: 28.03 / Avg: 274.61 / Max: 310.69 Min: 43.52 / Avg: 282.96 / Max: 321.53 Min: 71.44 / Avg: 288.1 / Max: 318.09 Min: 22.46 / Avg: 289.87 / Max: 333.39 Min: 23.7 / Avg: 308.68 / Max: 376.19 Min: 54.97 / Avg: 308.91 / Max: 345.11 Min: 32.66 / Avg: 362.61 / Max: 410.64 Min: 108.45 / Avg: 444.03 / Max: 499.02 Min: 107.82 / Avg: 445 / Max: 497.2 Min: 99.96 / Avg: 488.8 / Max: 580.11 Min: 183.06 / Avg: 553.66 / Max: 618.19 Min: 54.52 / Avg: 556.51 / Max: 637.38 Min: 109.84 / Avg: 561.7 / Max: 638.73 Min: 42.7 / Avg: 583.61 / Max: 678.24 Min: 186.19 / Avg: 605.28 / Max: 739.27 Min: 71.27 / Avg: 618.12 / Max: 752.32 Min: 168.16 / Avg: 627.31 / Max: 720.15 Min: 67.44 / Avg: 709.94 / Max: 820.99
Liquid-DSP LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 256 - Buffer Length: 256 - Filter Length: 57 EPYC 9684X 2P Xeon 6780E 2P EPYC 9654 2P Xeon 6766E 2P EPYC 9754 EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9684X Xeon Platinum 8490H 2P EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 EPYC 8534P Xeon Platinum 8380 2P Xeon Max 9468 EPYC 8534PN Xeon Platinum 8380 2000M 4000M 6000M 8000M 10000M SE +/- 50144890.07, N = 3 SE +/- 30821331.00, N = 3 SE +/- 60380193.59, N = 3 SE +/- 28436830.58, N = 3 SE +/- 12739745.33, N = 3 SE +/- 8069352.17, N = 3 SE +/- 61971606.40, N = 3 SE +/- 8144527.82, N = 3 SE +/- 2331189.49, N = 3 SE +/- 2618735.91, N = 3 SE +/- 28453880.81, N = 3 SE +/- 17558125.69, N = 3 SE +/- 6906036.01, N = 3 SE +/- 8015887.00, N = 3 SE +/- 3555434.03, N = 3 SE +/- 5292237.50, N = 3 SE +/- 1976810.00, N = 3 SE +/- 1959875.28, N = 3 SE +/- 3572580.77, N = 3 SE +/- 4088330.28, N = 3 SE +/- 4481443.22, N = 3 SE +/- 2935605.18, N = 3 8012700000 7597566667 7411866667 6800900000 6401366667 6316266667 6284300000 5811900000 5361266667 5266466667 5036000000 4825133333 4714700000 4259433333 3806833333 3500833333 3222533333 3116933333 3037100000 2848933333 2793700000 1610466667 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 256 - Buffer Length: 256 - Filter Length: 57 EPYC 8534PN EPYC 9754 EPYC 8534P Xeon 6766E EPYC 9654 EPYC 9754 2P EPYC 9684X Xeon 6780E Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8592+ EPYC 9684X 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8592+ 2P Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 5M 10M 15M 20M 25M 23457026.73 22424548.82 21345877.02 18991058.67 17475938.77 16347865.87 16030850.56 15942975.31 15666459.94 13780225.29 12864302.70 12268330.08 11934150.92 11085612.93 10156047.28 10144378.39 8954185.90 8419315.63 7871904.97 7597124.43 6601048.41 6213581.60 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X EPYC 9754 2P Xeon 6766E 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9684X 2P 140 280 420 560 700 Min: 18.06 / Avg: 119.1 / Max: 135.86 Min: 17.88 / Avg: 146.02 / Max: 166.38 Min: 43.61 / Avg: 224.29 / Max: 267.77 Min: 52.71 / Avg: 243.97 / Max: 290.99 Min: 22.44 / Avg: 285.46 / Max: 328.97 Min: 43.65 / Avg: 295.72 / Max: 355.96 Min: 30.72 / Avg: 301.36 / Max: 345.8 Min: 24.86 / Avg: 310.3 / Max: 377.83 Min: 58.98 / Avg: 315.8 / Max: 377.13 Min: 75.49 / Avg: 317.3 / Max: 376.02 Min: 75.93 / Avg: 318.17 / Max: 376.27 Min: 35.02 / Avg: 362.54 / Max: 411.45 Min: 40.8 / Avg: 386.37 / Max: 445.76 Min: 106.08 / Avg: 434.11 / Max: 495.15 Min: 101.61 / Avg: 488.78 / Max: 580.58 Min: 108.43 / Avg: 551.34 / Max: 625.76 Min: 54.92 / Avg: 576.16 / Max: 661.98 Min: 68.73 / Avg: 619.49 / Max: 754.72 Min: 195.33 / Avg: 635.13 / Max: 752.27 Min: 155.13 / Avg: 636.78 / Max: 752.99 Min: 193.94 / Avg: 639.74 / Max: 752.86 Min: 62.88 / Avg: 671.41 / Max: 791.35
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 256 - Buffer Length: 256 - Filter Length: 32 EPYC 9684X 2P Xeon 6780E 2P EPYC 9654 2P Xeon 6766E 2P EPYC 9754 EPYC 9754 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P EPYC 9684X Xeon Max 9480 2P EPYC 9654 Xeon 6780E Xeon Max 9468 2P Xeon 6766E Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Platinum 8592+ EPYC 8534P EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2000M 4000M 6000M 8000M 10000M SE +/- 34610274.65, N = 3 SE +/- 20501246.57, N = 3 SE +/- 46730658.51, N = 3 SE +/- 32054242.22, N = 3 SE +/- 3768731.00, N = 3 SE +/- 2936551.27, N = 3 SE +/- 10496401.50, N = 3 SE +/- 6711433.03, N = 3 SE +/- 11426625.83, N = 3 SE +/- 895048.11, N = 3 SE +/- 1419311.26, N = 3 SE +/- 2601495.30, N = 3 SE +/- 2447674.63, N = 3 SE +/- 821245.67, N = 3 SE +/- 7369833.41, N = 3 SE +/- 4021608.30, N = 3 SE +/- 2516611.48, N = 3 SE +/- 1501110.70, N = 3 SE +/- 1928730.15, N = 3 SE +/- 1338323.99, N = 3 SE +/- 1320353.49, N = 3 SE +/- 1266666.67, N = 3 7935266667 7546366667 7513333333 6821533333 6588200000 6439000000 6116566667 6077300000 5862466667 5367033333 5271933333 4799133333 4670966667 4301166667 3572233333 3243800000 3167200000 3062200000 2789600000 2769066667 2381800000 1823566667 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 256 - Buffer Length: 256 - Filter Length: 32 EPYC 8534PN EPYC 9754 EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9654 Xeon 6766E 2P EPYC 9754 2P EPYC 9684X Xeon 6780E 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9468 Xeon Max 9480 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Max 9468 2P 5M 10M 15M 20M 25M 23304528.15 22727990.09 21992105.60 21409914.84 18436155.38 17844515.21 17638157.23 16713460.67 16171511.91 15839180.07 13431732.69 12478169.67 10249944.70 10180344.31 9752693.41 9661394.05 9061082.03 8576477.66 8335236.32 7475059.61 7299110.00 7263532.04 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon Platinum 8380 Xeon 6780E Xeon Max 9468 EPYC 9754 EPYC 9654 Xeon Max 9480 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9684X EPYC 9754 2P Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 16.68 / Avg: 119.7 / Max: 139.58 Min: 18.82 / Avg: 139.24 / Max: 159.66 Min: 43.68 / Avg: 200.9 / Max: 230.57 Min: 52.43 / Avg: 243.95 / Max: 291.4 Min: 43.63 / Avg: 260.31 / Max: 300.33 Min: 74.86 / Avg: 277.71 / Max: 312.01 Min: 21.58 / Avg: 289.87 / Max: 330.19 Min: 30.54 / Avg: 295.44 / Max: 336.72 Min: 78.49 / Avg: 305.6 / Max: 341.63 Min: 25.92 / Avg: 311.11 / Max: 357.53 Min: 59.61 / Avg: 316.47 / Max: 377.01 Min: 34.34 / Avg: 362.52 / Max: 411.45 Min: 40.61 / Avg: 385.26 / Max: 440.96 Min: 104.79 / Avg: 386.75 / Max: 438.06 Min: 107.72 / Avg: 476.44 / Max: 538.02 Min: 101.34 / Avg: 489.41 / Max: 581.58 Min: 59.09 / Avg: 559.37 / Max: 644.46 Min: 70.49 / Avg: 623.14 / Max: 733.57 Min: 160.12 / Avg: 633.09 / Max: 754.38 Min: 65.54 / Avg: 635.93 / Max: 742.68 Min: 186.6 / Avg: 643.07 / Max: 754.55 Min: 204.52 / Avg: 643.9 / Max: 752.56
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 128 - Buffer Length: 256 - Filter Length: 32 EPYC 9684X 2P EPYC 9684X Xeon 6780E EPYC 9654 2P EPYC 9654 Xeon 6780E 2P EPYC 9754 2P EPYC 9754 Xeon 6766E Xeon 6766E 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H Xeon Platinum 8592+ EPYC 8534P Xeon Platinum 8380 2P EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 900M 1800M 2700M 3600M 4500M SE +/- 4250882.26, N = 3 SE +/- 841295.30, N = 3 SE +/- 2023198.79, N = 3 SE +/- 3186603.90, N = 3 SE +/- 2714365.08, N = 3 SE +/- 517472.49, N = 3 SE +/- 384418.75, N = 3 SE +/- 4056818.68, N = 3 SE +/- 920748.49, N = 3 SE +/- 3194265.70, N = 3 SE +/- 26633124.74, N = 3 SE +/- 13813198.20, N = 3 SE +/- 2403700.85, N = 3 SE +/- 4598671.31, N = 3 SE +/- 1858314.65, N = 3 SE +/- 2236316.42, N = 3 SE +/- 3138116.92, N = 3 SE +/- 3557152.42, N = 3 SE +/- 2003607.86, N = 3 SE +/- 1342054.81, N = 3 SE +/- 400000.00, N = 3 SE +/- 3178574.10, N = 3 4407900000 4353533333 4102200000 4063333333 4032733333 4021033333 3794133333 3791566667 3671966667 3607900000 3558600000 3520766667 3337066667 3254033333 3200600000 3148233333 3047066667 3044700000 2767933333 2739633333 2351900000 1797400000 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 128 - Buffer Length: 256 - Filter Length: 32 EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon 6766E 2P Xeon 6780E 2P EPYC 9684X EPYC 9754 2P Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9654 2P Xeon Max 9480 EPYC 9684X 2P Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Max 9468 2P 5M 10M 15M 20M 25M 23574098.60 21706213.46 20349128.58 17724724.88 15541749.24 14392112.20 14282136.05 13659351.74 11993608.96 10891038.63 10144782.58 10135010.79 9803609.24 8871933.76 8852490.97 8304651.14 7357841.69 6918246.02 6353287.78 6260399.40 5205818.86 5098423.00 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 Xeon 6766E 2P EPYC 9654 Xeon Max 9468 Xeon 6780E 2P Xeon Max 9480 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P EPYC 9684X EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 16.54 / Avg: 117.41 / Max: 138.28 Min: 16.53 / Avg: 140.38 / Max: 157.26 Min: 43.66 / Avg: 180.45 / Max: 204.69 Min: 43.44 / Avg: 231.44 / Max: 260.99 Min: 21.67 / Avg: 243.96 / Max: 276.65 Min: 52.85 / Avg: 244.28 / Max: 291.62 Min: 104.22 / Avg: 252.62 / Max: 279.84 Min: 28.95 / Avg: 280.2 / Max: 318.02 Min: 74.4 / Avg: 283.2 / Max: 376.87 Min: 107.24 / Avg: 294.38 / Max: 328.07 Min: 80.13 / Avg: 308.8 / Max: 366.56 Min: 27.32 / Avg: 310.33 / Max: 359.1 Min: 62.65 / Avg: 315.8 / Max: 377.95 Min: 40.25 / Avg: 348.37 / Max: 393.82 Min: 35.34 / Avg: 362.99 / Max: 409.42 Min: 56.36 / Avg: 414.47 / Max: 483.6 Min: 104.18 / Avg: 486.34 / Max: 574.85 Min: 66.83 / Avg: 497.93 / Max: 572.66 Min: 144.18 / Avg: 514.38 / Max: 660.61 Min: 75.15 / Avg: 554.16 / Max: 686.41 Min: 189.62 / Avg: 638.24 / Max: 714.51 Min: 196.59 / Avg: 641.03 / Max: 736.11
Google SynthMark SynthMark is a cross platform tool for benchmarking CPU performance under a variety of real-time audio workloads. It uses a polyphonic synthesizer model to provide standardized tests for latency, jitter and computational throughput. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Voices, More Is Better Google SynthMark 20201109 Test: VoiceMark_100 Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9654 EPYC 9684X EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9480 Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9468 2P EPYC 8534PN EPYC 9754 EPYC 8534P EPYC 9754 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E 2P Xeon 6766E 200 400 600 800 1000 SE +/- 0.13, N = 3 SE +/- 0.38, N = 3 SE +/- 0.55, N = 3 SE +/- 0.69, N = 3 SE +/- 0.72, N = 3 SE +/- 0.56, N = 3 SE +/- 1.51, N = 3 SE +/- 0.48, N = 3 SE +/- 0.24, N = 3 SE +/- 0.22, N = 3 SE +/- 0.26, N = 3 SE +/- 0.30, N = 3 SE +/- 0.60, N = 3 SE +/- 0.45, N = 3 SE +/- 0.14, N = 3 SE +/- 0.88, N = 3 SE +/- 7.09, N = 3 SE +/- 0.53, N = 3 SE +/- 0.18, N = 3 SE +/- 0.06, N = 3 SE +/- 0.33, N = 3 SE +/- 0.58, N = 3 799.72 799.49 783.78 783.42 783.20 782.32 712.70 712.09 712.01 711.38 711.10 710.92 656.25 655.80 655.24 655.05 555.60 548.36 540.24 538.91 494.38 494.05 1. (CXX) g++ options: -lm -lpthread -std=c++11 -Ofast Voices Per Watt
OpenBenchmarking.org Voices Per Watt, More Is Better Google SynthMark 20201109 Test: VoiceMark_100 EPYC 8534PN EPYC 8534P Xeon 6780E Xeon 6766E EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9684X Xeon Max 9468 Xeon Platinum 8380 Xeon Max 9480 Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Max 9468 2P 3 6 9 12 15 13.295 12.984 12.340 11.291 9.006 8.699 7.777 7.626 7.582 6.250 5.954 5.891 4.969 4.640 3.610 3.607 3.246 3.121 3.044 2.986 2.567 2.322 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Google SynthMark 20201109 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8490H Xeon Platinum 8592+ EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Max 9468 2P 70 140 210 280 350 Min: 43.51 / Avg: 43.76 / Max: 45.51 Min: 43.34 / Avg: 43.78 / Max: 45.89 Min: 16.18 / Avg: 49.36 / Max: 54.66 Min: 17.96 / Avg: 50.47 / Max: 55.88 Min: 21.02 / Avg: 72.82 / Max: 80.06 Min: 28.69 / Avg: 90.1 / Max: 99.51 Min: 52.55 / Avg: 93.31 / Max: 100.6 Min: 63.55 / Avg: 93.65 / Max: 130.75 Min: 33.54 / Avg: 102.83 / Max: 128.91 Min: 33.97 / Avg: 103.33 / Max: 113.89 Min: 104.37 / Avg: 106.55 / Max: 109.25 Min: 106.8 / Avg: 108.46 / Max: 112.7 Min: 75.06 / Avg: 113.78 / Max: 122.61 Min: 84.65 / Avg: 120.87 / Max: 129.59 Min: 40.35 / Avg: 181.43 / Max: 200.97 Min: 99.91 / Avg: 183.65 / Max: 205.87 Min: 57.99 / Avg: 216.87 / Max: 238.45 Min: 172.36 / Avg: 234.1 / Max: 329.35 Min: 61.62 / Avg: 241.28 / Max: 266.15 Min: 77.31 / Avg: 256.19 / Max: 292.81 Min: 201.92 / Avg: 277.45 / Max: 348.85 Min: 203.32 / Avg: 306.14 / Max: 413.85
Liquid-DSP LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 1.6 Threads: 64 - Buffer Length: 256 - Filter Length: 32 EPYC 9684X EPYC 9654 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon 6780E Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 9754 EPYC 9754 2P Xeon Platinum 8490H EPYC 8534P Xeon 6766E Xeon 6766E 2P EPYC 8534PN Xeon Platinum 8380 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 500M 1000M 1500M 2000M 2500M SE +/- 3000185.18, N = 3 SE +/- 2810891.52, N = 3 SE +/- 1567730.14, N = 3 SE +/- 2510644.01, N = 3 SE +/- 17236814.61, N = 3 SE +/- 3017909.50, N = 3 SE +/- 808290.38, N = 3 SE +/- 3300505.01, N = 3 SE +/- 4028647.42, N = 3 SE +/- 2206304.10, N = 3 SE +/- 27191440.65, N = 3 SE +/- 1550627.11, N = 3 SE +/- 1462114.15, N = 3 SE +/- 3167192.94, N = 3 SE +/- 3090487.20, N = 3 SE +/- 1379613.47, N = 3 SE +/- 1003881.36, N = 3 SE +/- 3256958.36, N = 3 SE +/- 9052685.30, N = 3 SE +/- 692820.32, N = 3 SE +/- 688799.28, N = 3 SE +/- 3209534.41, N = 3 2287466667 2282766667 2282066667 2279300000 2154633333 2090466667 2057300000 2025000000 1989300000 1963233333 1927766667 1898766667 1894966667 1890566667 1871333333 1841400000 1780533333 1776866667 1731166667 1679700000 1640266667 1561766667 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid samples/s Per Watt
OpenBenchmarking.org samples/s Per Watt, More Is Better Liquid-DSP 1.6 Threads: 64 - Buffer Length: 256 - Filter Length: 32 EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 EPYC 9654 Xeon 6780E 2P Xeon 6766E 2P EPYC 9684X EPYC 9754 2P Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 4M 8M 12M 16M 20M 17627091.46 15528502.68 14706486.76 14358014.80 11984070.91 9937920.60 9419289.56 9390076.73 8884551.10 7114576.18 6722664.60 6720931.25 6685402.04 6609424.93 6395094.29 6304024.25 5779445.49 4527054.56 4217240.97 3966847.18 3387722.73 3308597.89 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Liquid-DSP 1.6 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 EPYC 9684X EPYC 9754 2P Xeon Platinum 8490H Xeon Platinum 8592+ EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 120 240 360 480 600 Min: 15.81 / Avg: 100.8 / Max: 114.74 Min: 43.61 / Avg: 118.58 / Max: 131.69 Min: 15.76 / Avg: 127.25 / Max: 146.72 Min: 43.43 / Avg: 143.29 / Max: 159.65 Min: 20.46 / Avg: 158.44 / Max: 178.44 Min: 104.03 / Avg: 189.62 / Max: 209.49 Min: 105.02 / Avg: 211.19 / Max: 235.86 Min: 25.11 / Avg: 229.7 / Max: 257.97 Min: 49.48 / Avg: 244.21 / Max: 290.74 Min: 76.61 / Avg: 245.35 / Max: 268.44 Min: 74.71 / Avg: 254.14 / Max: 326.8 Min: 33.54 / Avg: 257.47 / Max: 292.35 Min: 38.6 / Avg: 266.35 / Max: 296.68 Min: 60.85 / Avg: 281.22 / Max: 356.43 Min: 27.76 / Avg: 286.83 / Max: 364.52 Min: 57.58 / Avg: 362 / Max: 403.93 Min: 64.31 / Avg: 394.38 / Max: 445.2 Min: 100.23 / Avg: 410.5 / Max: 516.01 Min: 75.17 / Avg: 475.95 / Max: 586.43 Min: 152.44 / Avg: 510.48 / Max: 610.23 Min: 200.67 / Avg: 593.37 / Max: 651.17 Min: 197.41 / Avg: 617.07 / Max: 678.02
uvg266 uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better uvg266 0.4.1 Video Input: Bosphorus 4K - Video Preset: Very Fast EPYC 9684X EPYC 9654 EPYC 9684X 2P Xeon Platinum 8490H 2P EPYC 9754 Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 8534P Xeon Platinum 8490H Xeon Max 9480 2P EPYC 9754 2P EPYC 8534PN Xeon Max 9468 2P Xeon 6780E Xeon Max 9480 Xeon 6766E Xeon Platinum 8380 2P Xeon Max 9468 Xeon Platinum 8592+ Xeon Platinum 8380 Xeon 6780E 2P Xeon 6766E 2P 14 28 42 56 70 SE +/- 0.13, N = 5 SE +/- 0.13, N = 5 SE +/- 0.12, N = 5 SE +/- 0.50, N = 5 SE +/- 0.18, N = 5 SE +/- 0.16, N = 5 SE +/- 0.55, N = 5 SE +/- 0.08, N = 4 SE +/- 0.01, N = 4 SE +/- 0.38, N = 9 SE +/- 0.43, N = 15 SE +/- 0.06, N = 4 SE +/- 0.39, N = 4 SE +/- 0.06, N = 4 SE +/- 0.02, N = 4 SE +/- 0.01, N = 4 SE +/- 0.18, N = 4 SE +/- 0.12, N = 4 SE +/- 0.08, N = 3 SE +/- 0.06, N = 3 SE +/- 0.43, N = 15 SE +/- 0.34, N = 15 62.70 62.02 55.05 54.88 54.46 53.66 53.44 50.04 49.42 46.25 46.00 44.92 44.75 44.44 40.82 40.73 40.69 39.56 35.10 30.17 28.43 27.33 Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better uvg266 0.4.1 Video Input: Bosphorus 4K - Video Preset: Very Fast EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8490H EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ Xeon 6766E 2P EPYC 9684X 2P Xeon 6780E 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.1087 0.2174 0.3261 0.4348 0.5435 0.483 0.445 0.389 0.369 0.346 0.310 0.269 0.193 0.185 0.167 0.160 0.157 0.154 0.152 0.150 0.147 0.138 0.132 0.121 0.110 0.092 0.090 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better uvg266 0.4.1 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9684X EPYC 9754 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 110 220 330 440 550 Min: 17.16 / Avg: 92.93 / Max: 127.1 Min: 43.61 / Avg: 104.71 / Max: 124.97 Min: 17.76 / Avg: 112.52 / Max: 149.73 Min: 43.47 / Avg: 120.53 / Max: 150.37 Min: 18.78 / Avg: 157.17 / Max: 216.69 Min: 104.1 / Avg: 174.26 / Max: 196.44 Min: 106.09 / Avg: 187.3 / Max: 229.82 Min: 28.57 / Avg: 200.36 / Max: 279.14 Min: 53.16 / Avg: 217.91 / Max: 258.49 Min: 27.1 / Avg: 218.83 / Max: 315.11 Min: 33.4 / Avg: 232.88 / Max: 324.16 Min: 39.54 / Avg: 249.19 / Max: 335.92 Min: 59.83 / Avg: 256.47 / Max: 330.75 Min: 76.83 / Avg: 268.38 / Max: 326.31 Min: 73.99 / Avg: 272.27 / Max: 334.62 Min: 52.01 / Avg: 320.69 / Max: 430.96 Min: 57.76 / Avg: 357.6 / Max: 478.64 Min: 103.41 / Avg: 370.67 / Max: 458.86 Min: 77.02 / Avg: 407.69 / Max: 581.55 Min: 161.1 / Avg: 454.5 / Max: 579.21 Min: 193.43 / Avg: 495.22 / Max: 603.09 Min: 191.29 / Avg: 503.31 / Max: 608.25
Blender Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.1 Blend File: Junkshop - Compute: CPU-Only EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P EPYC 9754 Xeon Platinum 8592+ 2P EPYC 9684X Xeon 6766E 2P EPYC 9654 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E Xeon Platinum 8592+ Xeon Platinum 8380 2P EPYC 8534P Xeon Platinum 8490H EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 14 28 42 56 70 SE +/- 0.02, N = 4 SE +/- 0.08, N = 4 SE +/- 0.02, N = 4 SE +/- 0.14, N = 3 SE +/- 0.08, N = 3 SE +/- 0.09, N = 3 SE +/- 0.13, N = 3 SE +/- 0.11, N = 3 SE +/- 0.06, N = 3 SE +/- 0.04, N = 3 SE +/- 0.19, N = 3 SE +/- 0.06, N = 3 SE +/- 0.03, N = 3 SE +/- 0.08, N = 3 SE +/- 0.10, N = 3 SE +/- 0.10, N = 3 SE +/- 0.06, N = 3 SE +/- 0.24, N = 3 SE +/- 0.03, N = 3 SE +/- 0.10, N = 3 SE +/- 0.04, N = 3 SE +/- 0.09, N = 3 11.21 11.47 12.87 16.89 17.70 18.31 18.65 18.70 21.02 21.15 23.45 24.99 26.49 29.06 32.08 33.20 35.82 35.96 38.86 39.17 42.85 61.22 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Blender 4.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9684X Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 18.29 / Avg: 127.16 / Max: 145.93 Min: 18.72 / Avg: 148.28 / Max: 170.5 Min: 43.7 / Avg: 186.05 / Max: 225.88 Min: 43.63 / Avg: 235.7 / Max: 293.05 Min: 51.88 / Avg: 251.6 / Max: 290.8 Min: 22.09 / Avg: 255.66 / Max: 340.71 Min: 16.01 / Avg: 272.23 / Max: 337.39 Min: 24.83 / Avg: 303.23 / Max: 378.56 Min: 59.14 / Avg: 311.52 / Max: 376.21 Min: 35.12 / Avg: 316.57 / Max: 411.05 Min: 85.92 / Avg: 320.19 / Max: 376.39 Min: 75.79 / Avg: 322.02 / Max: 374.53 Min: 106.82 / Avg: 347.76 / Max: 500.93 Min: 107.64 / Avg: 397.69 / Max: 593.53 Min: 39.86 / Avg: 438.61 / Max: 691.32 Min: 105.08 / Avg: 471.81 / Max: 581.38 Min: 59.75 / Avg: 483.45 / Max: 681.8 Min: 77.61 / Avg: 545.55 / Max: 722.02 Min: 63.79 / Avg: 553.98 / Max: 822.05 Min: 169.62 / Avg: 575.47 / Max: 754.28 Min: 206.83 / Avg: 603.68 / Max: 753.22 Min: 208.58 / Avg: 605.47 / Max: 751.55
Result
OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.1 Blend File: Fishy Cat - Compute: CPU-Only EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 Xeon Platinum 8592+ 2P EPYC 9684X Xeon Platinum 8490H 2P EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon 6766E Xeon Platinum 8380 2P Xeon Platinum 8592+ EPYC 8534P Xeon Platinum 8490H EPYC 8534PN Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 13 26 39 52 65 SE +/- 0.04, N = 5 SE +/- 0.02, N = 5 SE +/- 0.05, N = 4 SE +/- 0.01, N = 4 SE +/- 0.05, N = 3 SE +/- 0.13, N = 3 SE +/- 0.04, N = 3 SE +/- 0.07, N = 3 SE +/- 0.08, N = 3 SE +/- 0.11, N = 3 SE +/- 0.18, N = 3 SE +/- 0.13, N = 3 SE +/- 0.08, N = 3 SE +/- 0.37, N = 3 SE +/- 0.27, N = 3 SE +/- 0.06, N = 3 SE +/- 0.23, N = 3 SE +/- 0.11, N = 3 SE +/- 0.07, N = 3 SE +/- 0.04, N = 3 SE +/- 0.17, N = 3 SE +/- 0.15, N = 3 9.71 10.16 11.25 15.16 16.52 16.58 17.29 18.36 19.13 20.15 21.45 23.17 26.52 29.55 30.75 31.91 34.35 34.83 38.04 38.48 42.78 59.15 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Blender 4.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E Xeon Platinum 8380 EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 16.69 / Avg: 126.35 / Max: 143.05 Min: 16.84 / Avg: 152.85 / Max: 176.56 Min: 43.64 / Avg: 193.48 / Max: 236.22 Min: 43.69 / Avg: 247.1 / Max: 308.06 Min: 53.1 / Avg: 251.93 / Max: 290.63 Min: 22.97 / Avg: 261.77 / Max: 349.11 Min: 30.77 / Avg: 277.68 / Max: 346.38 Min: 23.46 / Avg: 303.18 / Max: 371.62 Min: 59.73 / Avg: 313.33 / Max: 377.56 Min: 74.14 / Avg: 319.1 / Max: 376.73 Min: 74.34 / Avg: 321.04 / Max: 375.24 Min: 36.44 / Avg: 322.13 / Max: 410.23 Min: 106.69 / Avg: 350.36 / Max: 494.32 Min: 109.2 / Avg: 429.78 / Max: 624.43 Min: 42.38 / Avg: 447.74 / Max: 705.75 Min: 104.59 / Avg: 473.37 / Max: 583.21 Min: 60.34 / Avg: 491.93 / Max: 710.82 Min: 76.09 / Avg: 546.98 / Max: 735.92 Min: 67.56 / Avg: 561.28 / Max: 821.67 Min: 169.36 / Avg: 570.45 / Max: 755.96 Min: 211.08 / Avg: 602.46 / Max: 751.81 Min: 205.55 / Avg: 607.36 / Max: 752.78
Appleseed Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.
Scene: Disney Material
EPYC 9684X 2P: The test run did not produce a result.
EPYC 9654 2P: The test run did not produce a result.
EPYC 9754 2P: The test run did not produce a result.
Xeon 6766E 2P: The test run did not produce a result.
Xeon 6780E 2P: The test run did not produce a result.
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 64 - Model: AlexNet Xeon Platinum 8490H 2P Xeon Platinum 8490H EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9754 Xeon Platinum 8380 2P EPYC 9684X Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8592+ EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8380 EPYC 8534P EPYC 9754 2P EPYC 8534PN Xeon 6780E 2P Xeon 6780E Xeon 6766E 2P Xeon 6766E 200 400 600 800 1000 SE +/- 7.98, N = 5 SE +/- 0.51, N = 5 SE +/- 1.31, N = 5 SE +/- 7.35, N = 5 SE +/- 6.39, N = 13 SE +/- 2.99, N = 5 SE +/- 3.70, N = 5 SE +/- 1.37, N = 5 SE +/- 2.22, N = 5 SE +/- 2.31, N = 5 SE +/- 1.95, N = 5 SE +/- 6.22, N = 5 SE +/- 7.66, N = 5 SE +/- 8.87, N = 4 SE +/- 0.27, N = 4 SE +/- 0.45, N = 4 SE +/- 7.43, N = 15 SE +/- 0.16, N = 4 SE +/- 5.91, N = 15 SE +/- 0.97, N = 3 SE +/- 5.44, N = 15 SE +/- 0.70, N = 3 996.47 969.62 942.19 940.40 932.59 924.80 904.36 868.54 858.84 852.36 841.68 819.89 738.31 727.77 690.09 673.42 658.97 617.03 520.01 482.06 464.80 443.74 images/sec Per Watt
OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 64 - Model: AlexNet EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8490H EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Max 9480 Xeon 6766E Xeon Max 9468 Xeon 6780E Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9754 2P EPYC 9654 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9684X 2P Xeon 6780E 2P Xeon 6766E 2P 2 4 6 8 10 6.664 6.508 6.238 5.439 4.375 4.332 4.145 3.481 3.374 3.355 3.341 3.051 2.677 2.404 2.269 2.236 2.065 1.919 1.898 1.857 1.794 1.740 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.16.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 120 240 360 480 600 Min: 15.55 / Avg: 92.59 / Max: 128.11 Min: 17.62 / Avg: 103.47 / Max: 147.02 Min: 43.64 / Avg: 132.25 / Max: 173.77 Min: 21.07 / Avg: 148.26 / Max: 241.29 Min: 43.45 / Avg: 158.02 / Max: 214.1 Min: 27.88 / Avg: 173.22 / Max: 270.57 Min: 33.53 / Avg: 197.79 / Max: 292.33 Min: 53.17 / Avg: 198.24 / Max: 276.92 Min: 32.91 / Avg: 200.48 / Max: 302.35 Min: 58.85 / Avg: 221.64 / Max: 332.07 Min: 72.44 / Avg: 251.93 / Max: 351.73 Min: 75.5 / Avg: 252.6 / Max: 351.73 Min: 104.81 / Avg: 267.11 / Max: 360.8 Min: 107.28 / Avg: 289.41 / Max: 402.43 Min: 40.23 / Avg: 294.77 / Max: 430.69 Min: 99.68 / Avg: 337.77 / Max: 509.02 Min: 59.14 / Avg: 352.38 / Max: 512.39 Min: 76.02 / Avg: 357.25 / Max: 522.13 Min: 64.28 / Avg: 397.64 / Max: 581.36 Min: 156.12 / Avg: 439.21 / Max: 636.86 Min: 186.67 / Avg: 489.99 / Max: 671.14 Min: 196.31 / Avg: 491.23 / Max: 676.71
Kvazaar This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better Kvazaar 2.2 Video Input: Bosphorus 4K - Video Preset: Medium EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 EPYC 9754 Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon Max 9480 2P Xeon 6766E 2P Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 8534P EPYC 8534PN Xeon 6780E Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon 6766E Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 14 28 42 56 70 SE +/- 0.40, N = 5 SE +/- 0.56, N = 5 SE +/- 0.52, N = 15 SE +/- 0.27, N = 4 SE +/- 0.01, N = 4 SE +/- 0.07, N = 4 SE +/- 0.04, N = 4 SE +/- 0.06, N = 3 SE +/- 0.23, N = 3 SE +/- 0.07, N = 3 SE +/- 0.21, N = 3 SE +/- 0.08, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.03, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.00, N = 3 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 63.48 61.11 54.57 41.58 40.50 39.58 39.08 32.04 31.70 29.62 29.15 26.69 24.80 24.31 23.02 21.39 20.97 20.33 19.34 18.84 18.13 15.51 1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Kvazaar 2.2 Video Input: Bosphorus 4K - Video Preset: Medium EPYC 8534PN EPYC 9754 EPYC 8534P Xeon 6766E EPYC 9654 Xeon 6780E EPYC 9754 2P EPYC 9654 2P EPYC 9684X Xeon 6766E 2P EPYC 9684X 2P Xeon 6780E 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Platinum 8380 2P Xeon Max 9468 2P 0.0522 0.1044 0.1566 0.2088 0.261 0.232 0.207 0.200 0.185 0.181 0.172 0.169 0.151 0.148 0.135 0.135 0.130 0.090 0.082 0.080 0.066 0.066 0.065 0.065 0.054 0.053 0.050 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Kvazaar 2.2 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 Xeon 6766E 2P EPYC 9654 Xeon Platinum 8380 Xeon 6780E 2P Xeon Platinum 8490H EPYC 9684X Xeon Platinum 8592+ Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 120 240 360 480 600 Min: 15.7 / Avg: 99.33 / Max: 124.48 Min: 43.62 / Avg: 104.69 / Max: 123.24 Min: 17.56 / Avg: 121.69 / Max: 146.12 Min: 43.55 / Avg: 124.03 / Max: 147.77 Min: 20.71 / Avg: 188.76 / Max: 248.4 Min: 104.63 / Avg: 216.52 / Max: 253.59 Min: 26.12 / Avg: 219.12 / Max: 294.65 Min: 53 / Avg: 234.37 / Max: 263.04 Min: 106.92 / Avg: 243.13 / Max: 293.72 Min: 61.73 / Avg: 263.01 / Max: 306.52 Min: 34.18 / Avg: 273.99 / Max: 354.7 Min: 33.8 / Avg: 276.47 / Max: 362.01 Min: 77.93 / Avg: 280.57 / Max: 315.37 Min: 69.51 / Avg: 286.2 / Max: 325.03 Min: 38.26 / Avg: 323.31 / Max: 486.48 Min: 103.66 / Avg: 385.29 / Max: 444.98 Min: 56.07 / Avg: 404.28 / Max: 599.6 Min: 66.41 / Avg: 469.54 / Max: 685.47 Min: 168.08 / Avg: 490.42 / Max: 593.69 Min: 75.75 / Avg: 509.74 / Max: 682.24 Min: 209.31 / Avg: 533.47 / Max: 617.43 Min: 189.77 / Avg: 550.48 / Max: 652.57
Pennant Pennant is an application focused on hydrodynamics on general unstructured meshes in 2D. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Hydro Cycle Time - Seconds, Fewer Is Better Pennant 1.0.1 Test: sedovbig EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P EPYC 9754 EPYC 9684X EPYC 9654 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 8534P Xeon Max 9480 2P EPYC 8534PN Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 6 12 18 24 30 SE +/- 0.037087, N = 15 SE +/- 0.041359, N = 15 SE +/- 0.040626, N = 15 SE +/- 0.074320, N = 15 SE +/- 0.080453, N = 15 SE +/- 0.052975, N = 7 SE +/- 0.018471, N = 4 SE +/- 0.016400, N = 4 SE +/- 0.024479, N = 4 SE +/- 0.003253, N = 4 SE +/- 0.029749, N = 4 SE +/- 0.003237, N = 4 SE +/- 0.052953, N = 3 SE +/- 0.002987, N = 3 SE +/- 0.002400, N = 3 SE +/- 0.013353, N = 3 SE +/- 0.001847, N = 3 SE +/- 0.034794, N = 3 2.940456 3.081108 3.408337 5.147658 5.460660 5.916458 9.964526 11.284380 11.369640 12.106900 12.151320 13.211250 15.247190 16.030800 16.969100 19.882460 21.021120 25.789450 1. (CXX) g++ options: -fopenmp -lmpi_cxx -lmpi CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Pennant 1.0.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 16.13 / Avg: 110.49 / Max: 145.79 Min: 19.05 / Avg: 123.57 / Max: 166.99 Min: 20.25 / Avg: 182.99 / Max: 330.83 Min: 26.98 / Avg: 200.64 / Max: 329.21 Min: 51.49 / Avg: 239.83 / Max: 289.95 Min: 30.95 / Avg: 242.71 / Max: 409.53 Min: 26.41 / Avg: 279.92 / Max: 376.36 Min: 57.56 / Avg: 287.94 / Max: 376.04 Min: 78.91 / Avg: 298.37 / Max: 377.04 Min: 72.63 / Avg: 303.18 / Max: 377.15 Min: 41.72 / Avg: 310.27 / Max: 663.85 Min: 55.22 / Avg: 351.84 / Max: 674.58 Min: 64.82 / Avg: 404.01 / Max: 819.66 Min: 101.05 / Avg: 438.87 / Max: 580.29 Min: 69.95 / Avg: 489.66 / Max: 753.78 Min: 148.41 / Avg: 535.71 / Max: 753.56 Min: 206.49 / Avg: 569.16 / Max: 751.77 Min: 189.6 / Avg: 573.95 / Max: 754.13
Kvazaar This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better Kvazaar 2.2 Video Input: Bosphorus 4K - Video Preset: Slow EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 EPYC 9754 Xeon 6780E 2P Xeon Platinum 8490H 2P Xeon 6766E 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 8534P EPYC 8534PN Xeon 6780E Xeon Platinum 8490H Xeon Platinum 8380 2P Xeon 6766E Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 14 28 42 56 70 SE +/- 0.26, N = 5 SE +/- 0.50, N = 5 SE +/- 0.47, N = 8 SE +/- 0.04, N = 4 SE +/- 0.07, N = 4 SE +/- 0.01, N = 4 SE +/- 0.05, N = 4 SE +/- 0.18, N = 3 SE +/- 0.08, N = 3 SE +/- 0.09, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.07, N = 3 SE +/- 0.04, N = 3 SE +/- 0.05, N = 3 SE +/- 0.07, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.07, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.00, N = 3 62.82 61.07 54.88 40.92 40.00 38.90 38.47 31.85 31.24 29.15 28.96 26.00 24.23 23.71 22.46 20.87 20.23 19.71 18.86 18.24 17.52 15.09 1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Kvazaar 2.2 Video Input: Bosphorus 4K - Video Preset: Slow EPYC 8534PN EPYC 9754 EPYC 8534P Xeon 6766E EPYC 9754 2P EPYC 9654 Xeon 6780E EPYC 9654 2P EPYC 9684X EPYC 9684X 2P Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Platinum 8380 2P Xeon Max 9468 2P 0.0513 0.1026 0.1539 0.2052 0.2565 0.228 0.202 0.199 0.177 0.169 0.169 0.166 0.149 0.146 0.135 0.132 0.128 0.087 0.080 0.077 0.064 0.064 0.063 0.063 0.052 0.051 0.048 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Kvazaar 2.2 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 Xeon 6766E 2P EPYC 9654 Xeon Platinum 8380 Xeon 6780E 2P Xeon Platinum 8490H EPYC 9684X Xeon Platinum 8592+ Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9468 2P Xeon Max 9480 2P 120 240 360 480 600 Min: 15.1 / Avg: 98.52 / Max: 118.03 Min: 43.61 / Avg: 106.79 / Max: 125.47 Min: 17.7 / Avg: 119.03 / Max: 142.44 Min: 43.47 / Avg: 125.76 / Max: 153.18 Min: 21.44 / Avg: 190.72 / Max: 249.47 Min: 104.31 / Avg: 220.37 / Max: 264.43 Min: 28.16 / Avg: 229.8 / Max: 294.04 Min: 52.23 / Avg: 234.47 / Max: 262.38 Min: 105.65 / Avg: 248.74 / Max: 300.85 Min: 58.28 / Avg: 262.77 / Max: 303.74 Min: 30.18 / Avg: 273.17 / Max: 355.5 Min: 28.15 / Avg: 278.75 / Max: 351.38 Min: 80.03 / Avg: 279.05 / Max: 314.96 Min: 67.53 / Avg: 286.96 / Max: 322.9 Min: 38.94 / Avg: 325.22 / Max: 485.91 Min: 96.13 / Avg: 384.99 / Max: 445.25 Min: 58.04 / Avg: 409.96 / Max: 602.7 Min: 59.79 / Avg: 464.35 / Max: 680.95 Min: 168.67 / Avg: 496.87 / Max: 590.71 Min: 71.05 / Avg: 508.41 / Max: 687.83 Min: 210.38 / Avg: 537.83 / Max: 613.5 Min: 195.26 / Avg: 552.42 / Max: 652.44
srsRAN Project srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Mbps, More Is Better srsRAN Project 23.10.1-20240325 Test: PUSCH Processor Benchmark, Throughput Thread Xeon Max 9480 2P Xeon Max 9480 Xeon Max 9468 2P Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9654 EPYC 9684X EPYC 9684X 2P EPYC 8534PN EPYC 8534P EPYC 9754 2P EPYC 9754 Xeon 6780E 2P Xeon 6780E Xeon 6766E Xeon 6766E 2P 30 60 90 120 150 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.07, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 129.7 129.7 129.7 129.7 129.7 129.7 129.5 129.5 112.3 112.3 112.2 112.2 112.2 112.2 112.1 112.1 112.1 112.1 52.6 52.5 49.8 49.8 1. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -fno-trapping-math -fno-math-errno -ldl Mbps Per Watt
OpenBenchmarking.org Mbps Per Watt, More Is Better srsRAN Project 23.10.1-20240325 Test: PUSCH Processor Benchmark, Throughput Thread EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8592+ Xeon 6780E Xeon 6766E EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon 6780E 2P EPYC 9684X 2P Xeon 6766E 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.511 1.022 1.533 2.044 2.555 2.271 2.234 1.551 1.257 1.201 1.196 1.193 1.139 1.095 0.965 0.750 0.736 0.637 0.603 0.564 0.524 0.487 0.476 0.467 0.426 0.347 0.345 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better srsRAN Project 23.10.1-20240325 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 EPYC 9754 2P Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 70 140 210 280 350 Min: 43.55 / Avg: 43.73 / Max: 45.56 Min: 21.98 / Avg: 44.02 / Max: 45.61 Min: 16.95 / Avg: 49.36 / Max: 56.04 Min: 16.97 / Avg: 50.19 / Max: 57.02 Min: 21.9 / Avg: 72.27 / Max: 82.25 Min: 27.03 / Avg: 89.27 / Max: 101.62 Min: 52.1 / Avg: 93.49 / Max: 102.87 Min: 34.3 / Avg: 102.46 / Max: 115.98 Min: 103.85 / Avg: 106.63 / Max: 109.21 Min: 104.39 / Avg: 107.97 / Max: 110.48 Min: 34.52 / Avg: 108.29 / Max: 130.82 Min: 59.93 / Avg: 134.44 / Max: 153.53 Min: 72.46 / Avg: 172.93 / Max: 196.41 Min: 40.92 / Avg: 175.84 / Max: 200.95 Min: 77.25 / Avg: 176.34 / Max: 199.62 Min: 97.13 / Avg: 186.14 / Max: 224.52 Min: 58.78 / Avg: 213.93 / Max: 242.48 Min: 80.8 / Avg: 229.54 / Max: 284.84 Min: 61.91 / Avg: 235.89 / Max: 267.82 Min: 160.99 / Avg: 304.32 / Max: 339.38 Min: 206.09 / Avg: 373.48 / Max: 412.46 Min: 207.62 / Avg: 375.78 / Max: 415.17
Kvazaar This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better Kvazaar 2.2 Video Input: Bosphorus 4K - Video Preset: Very Fast EPYC 9684X EPYC 9654 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P EPYC 9754 2P EPYC 8534P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 8534PN Xeon Platinum 8490H Xeon 6780E Xeon Platinum 8380 2P Xeon 6766E Xeon Max 9480 Xeon Max 9468 Xeon 6780E 2P Xeon Platinum 8592+ Xeon 6766E 2P Xeon Platinum 8380 20 40 60 80 100 SE +/- 0.32, N = 6 SE +/- 0.62, N = 6 SE +/- 0.28, N = 6 SE +/- 0.71, N = 6 SE +/- 0.21, N = 5 SE +/- 0.45, N = 5 SE +/- 0.36, N = 5 SE +/- 0.63, N = 15 SE +/- 0.07, N = 5 SE +/- 0.29, N = 4 SE +/- 0.33, N = 4 SE +/- 0.07, N = 4 SE +/- 0.08, N = 4 SE +/- 0.00, N = 4 SE +/- 0.18, N = 4 SE +/- 0.04, N = 4 SE +/- 0.08, N = 4 SE +/- 0.06, N = 4 SE +/- 0.50, N = 15 SE +/- 0.04, N = 4 SE +/- 0.36, N = 15 SE +/- 0.06, N = 3 78.15 76.83 74.03 72.68 68.56 67.64 65.38 59.01 54.60 52.23 51.91 51.51 50.03 46.18 44.11 41.86 41.34 40.23 38.92 38.02 36.40 31.25 1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Kvazaar 2.2 Video Input: Bosphorus 4K - Video Preset: Very Fast EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6766E Xeon 6780E EPYC 9654 EPYC 9684X EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon 6766E 2P Xeon Platinum 8490H Xeon 6780E 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.1332 0.2664 0.3996 0.5328 0.666 0.592 0.507 0.470 0.449 0.429 0.406 0.359 0.249 0.241 0.219 0.213 0.212 0.210 0.183 0.175 0.163 0.162 0.156 0.152 0.129 0.109 0.109 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Kvazaar 2.2 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E Xeon 6780E EPYC 8534P EPYC 9754 Xeon 6766E 2P Xeon 6780E 2P EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 110 220 330 440 550 Min: 15.21 / Avg: 87.02 / Max: 118.81 Min: 43.56 / Avg: 93.24 / Max: 111.07 Min: 43.47 / Avg: 107.62 / Max: 131.78 Min: 17.32 / Avg: 107.72 / Max: 144.07 Min: 21.76 / Avg: 145.81 / Max: 208.59 Min: 104.11 / Avg: 170.72 / Max: 196.41 Min: 106.42 / Avg: 185.19 / Max: 224.67 Min: 28.16 / Avg: 189.32 / Max: 280.14 Min: 53.82 / Avg: 205.21 / Max: 244.22 Min: 28.31 / Avg: 207.87 / Max: 283.73 Min: 32 / Avg: 217.45 / Max: 322.84 Min: 60.42 / Avg: 235.64 / Max: 301.41 Min: 39.35 / Avg: 237.06 / Max: 332.11 Min: 74.97 / Avg: 248.42 / Max: 303.88 Min: 71.86 / Avg: 253.88 / Max: 311.53 Min: 58.8 / Avg: 302.06 / Max: 438.45 Min: 66.01 / Avg: 338.64 / Max: 489.89 Min: 100.56 / Avg: 342.63 / Max: 431.07 Min: 73.01 / Avg: 374.34 / Max: 560.15 Min: 162.16 / Avg: 433.41 / Max: 581.27 Min: 190.55 / Avg: 476.69 / Max: 591.36 Min: 187.97 / Avg: 480.56 / Max: 609.82
OpenFOAM OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Small Mesh Size - Mesh Time EPYC 9684X EPYC 9684X 2P EPYC 9654 2P EPYC 9654 Xeon Max 9480 2P Xeon Platinum 8592+ 2P EPYC 9754 2P EPYC 9754 Xeon Platinum 8490H Xeon Platinum 8592+ Xeon Max 9468 Xeon Max 9480 Xeon Max 9468 2P EPYC 8534P Xeon Platinum 8490H 2P EPYC 8534PN Xeon Platinum 8380 2P Xeon Platinum 8380 9 18 27 36 45 25.93 26.82 27.74 28.03 28.84 29.33 29.47 30.38 31.38 31.70 31.78 32.00 32.48 33.67 35.60 36.82 36.86 39.12 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm
SVT-AV1 This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.0 Encoder Mode: Preset 8 - Input: Bosphorus 4K EPYC 9654 EPYC 9684X 2P EPYC 9654 2P EPYC 9684X Xeon Platinum 8592+ EPYC 9754 EPYC 9754 2P Xeon Platinum 8490H Xeon Platinum 8592+ 2P EPYC 8534P Xeon Max 9480 Xeon Platinum 8490H 2P EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Platinum 8380 Xeon Max 9468 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 20 40 60 80 100 SE +/- 0.56, N = 5 SE +/- 0.63, N = 5 SE +/- 0.40, N = 5 SE +/- 0.42, N = 5 SE +/- 0.22, N = 5 SE +/- 0.38, N = 5 SE +/- 0.31, N = 4 SE +/- 0.20, N = 5 SE +/- 0.79, N = 5 SE +/- 0.11, N = 4 SE +/- 0.29, N = 4 SE +/- 0.46, N = 4 SE +/- 0.67, N = 4 SE +/- 0.26, N = 4 SE +/- 0.52, N = 7 SE +/- 0.44, N = 15 SE +/- 0.09, N = 4 SE +/- 0.58, N = 5 SE +/- 0.02, N = 4 SE +/- 0.34, N = 3 SE +/- 0.03, N = 4 SE +/- 0.35, N = 15 87.99 87.82 86.34 86.30 79.44 78.74 77.60 70.91 70.43 68.19 65.36 65.22 64.79 62.75 59.10 55.98 55.24 54.82 50.82 47.25 46.48 42.86 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better SVT-AV1 2.0 Encoder Mode: Preset 8 - Input: Bosphorus 4K EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon 6766E Xeon 6780E EPYC 9684X Xeon Platinum 8592+ EPYC 9754 2P Xeon Platinum 8490H EPYC 9654 2P Xeon Platinum 8380 EPYC 9684X 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.2147 0.4294 0.6441 0.8588 1.0735 0.954 0.879 0.737 0.653 0.595 0.587 0.570 0.439 0.388 0.362 0.349 0.336 0.321 0.306 0.299 0.286 0.278 0.227 0.224 0.179 0.126 0.124 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better SVT-AV1 2.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 100 200 300 400 500 Min: 14.98 / Avg: 67.93 / Max: 107.89 Min: 16.05 / Avg: 77.57 / Max: 129.87 Min: 43.66 / Avg: 78.14 / Max: 117.48 Min: 43.44 / Avg: 86.61 / Max: 139.64 Min: 21.17 / Avg: 106.8 / Max: 187.26 Min: 28.36 / Avg: 134.72 / Max: 238.02 Min: 55.92 / Avg: 143.16 / Max: 199.83 Min: 31.53 / Avg: 151.3 / Max: 266.73 Min: 105.44 / Avg: 154.41 / Max: 213.01 Min: 51.15 / Avg: 164.49 / Max: 247.62 Min: 27.5 / Avg: 181.13 / Max: 316.97 Min: 59.76 / Avg: 195.92 / Max: 293.59 Min: 39.92 / Avg: 200.17 / Max: 302.35 Min: 76.76 / Avg: 226.08 / Max: 319.38 Min: 73.57 / Avg: 228.42 / Max: 325.25 Min: 56.85 / Avg: 247.1 / Max: 379.6 Min: 98.92 / Avg: 263.75 / Max: 406.42 Min: 62.22 / Avg: 273.59 / Max: 425.17 Min: 70.26 / Avg: 309.78 / Max: 476.6 Min: 171.54 / Avg: 364.29 / Max: 488.45 Min: 193.1 / Avg: 442.35 / Max: 576.63 Min: 196.2 / Avg: 442.62 / Max: 580.63
Intel Open Image Denoise Open Image Denoise is a denoising library for ray-tracing and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 2.2 Run: RTLightmap.hdr.4096x4096 - Device: CPU-Only Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9754 2P Xeon 6780E 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ Xeon 6766E 2P EPYC 9754 Xeon Platinum 8380 2P EPYC 9654 Xeon Platinum 8490H EPYC 9684X Xeon Max 9480 Xeon Max 9468 Xeon 6780E Xeon 6766E EPYC 8534P Xeon Platinum 8380 EPYC 8534PN 0.54 1.08 1.62 2.16 2.7 SE +/- 0.01, N = 4 SE +/- 0.01, N = 4 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.00, N = 3 SE +/- 0.02, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.02, N = 4 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 2.40 2.18 1.74 1.72 1.71 1.68 1.67 1.64 1.58 1.51 1.50 1.48 1.31 1.29 1.28 1.14 1.08 1.07 0.98 0.86 0.86 0.80 Images / Sec Per Watt
OpenBenchmarking.org Images / Sec Per Watt, More Is Better Intel Open Image Denoise 2.2 Run: RTLightmap.hdr.4096x4096 - Device: CPU-Only EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon 6780E Xeon 6766E Xeon 6766E 2P EPYC 9754 2P EPYC 9684X Xeon Max 9480 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon 6780E 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Max 9468 0.0016 0.0032 0.0048 0.0064 0.008 0.007 0.006 0.006 0.005 0.005 0.004 0.004 0.004 0.004 0.004 0.004 0.004 0.004 0.004 0.003 0.003 0.003 0.003 0.003 0.003 0.003 0.003 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Intel Open Image Denoise 2.2 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon 6780E Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9684X Xeon 6766E 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 17.12 / Avg: 118.45 / Max: 132.79 Min: 17.84 / Avg: 136.47 / Max: 153.36 Min: 43.68 / Avg: 220.35 / Max: 250.91 Min: 22.02 / Avg: 239.18 / Max: 293.59 Min: 53.73 / Avg: 245.89 / Max: 283.33 Min: 26.15 / Avg: 266.61 / Max: 316.67 Min: 43.6 / Avg: 279.34 / Max: 327.38 Min: 35.29 / Avg: 290 / Max: 378.6 Min: 59.44 / Avg: 304.48 / Max: 364.78 Min: 78.93 / Avg: 313 / Max: 373.31 Min: 75.68 / Avg: 314.57 / Max: 370.97 Min: 35.65 / Avg: 329.68 / Max: 387.49 Min: 105.15 / Avg: 418.44 / Max: 529.24 Min: 40.88 / Avg: 433.56 / Max: 546.1 Min: 104.72 / Avg: 461.16 / Max: 559.61 Min: 108.06 / Avg: 500.21 / Max: 643.27 Min: 60.67 / Avg: 504.98 / Max: 617.56 Min: 79.35 / Avg: 534.18 / Max: 740.08 Min: 169.3 / Avg: 573.61 / Max: 745.77 Min: 66.51 / Avg: 585.9 / Max: 726.86 Min: 206.4 / Avg: 605.57 / Max: 749.65 Min: 198.59 / Avg: 608.01 / Max: 748.72
Blender Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.1 Blend File: BMW27 - Compute: CPU-Only EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8592+ 2P EPYC 9754 Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9480 2P EPYC 9654 Xeon Max 9468 2P Xeon 6780E Xeon 6766E Xeon Platinum 8380 2P Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 8534P Xeon Max 9480 EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 11 22 33 44 55 SE +/- 0.02, N = 6 SE +/- 0.03, N = 6 SE +/- 0.02, N = 5 SE +/- 0.09, N = 5 SE +/- 0.06, N = 5 SE +/- 0.03, N = 4 SE +/- 0.01, N = 4 SE +/- 0.08, N = 4 SE +/- 0.04, N = 4 SE +/- 0.08, N = 4 SE +/- 0.07, N = 4 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.17, N = 8 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.15, N = 3 SE +/- 0.04, N = 3 SE +/- 0.24, N = 3 SE +/- 0.05, N = 3 SE +/- 0.11, N = 3 SE +/- 0.11, N = 3 7.17 7.75 8.61 9.71 10.56 12.76 12.86 13.74 14.29 15.29 15.96 16.55 17.61 19.85 23.67 23.72 25.89 27.12 28.49 30.07 31.66 46.32 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Blender 4.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 EPYC 9684X Xeon Max 9468 Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 18.72 / Avg: 121.31 / Max: 138.75 Min: 17.13 / Avg: 146.16 / Max: 171.02 Min: 43.62 / Avg: 189.01 / Max: 235.56 Min: 43.52 / Avg: 237.13 / Max: 302.65 Min: 21.13 / Avg: 247.78 / Max: 338.02 Min: 52.7 / Avg: 250.52 / Max: 290.53 Min: 27.61 / Avg: 264.26 / Max: 339.47 Min: 33.51 / Avg: 297.61 / Max: 376.94 Min: 57.12 / Avg: 303.66 / Max: 376.88 Min: 80.66 / Avg: 311.66 / Max: 376.47 Min: 35.39 / Avg: 314.77 / Max: 411.05 Min: 72.63 / Avg: 316.78 / Max: 376.8 Min: 104.11 / Avg: 337.47 / Max: 486.69 Min: 106.78 / Avg: 401.76 / Max: 612.07 Min: 41.39 / Avg: 414.63 / Max: 680.91 Min: 56.75 / Avg: 465.75 / Max: 695.98 Min: 103.59 / Avg: 467.53 / Max: 580.29 Min: 66.83 / Avg: 521.92 / Max: 820.17 Min: 73.16 / Avg: 522.14 / Max: 732.74 Min: 171.9 / Avg: 556.95 / Max: 754.1 Min: 211.29 / Avg: 584.52 / Max: 755.18 Min: 204.3 / Avg: 589.35 / Max: 754.59
miniBUDE MiniBUDE is a mini application for the the core computation of the Bristol University Docking Engine (BUDE). This test profile currently makes use of the OpenMP implementation of miniBUDE for CPU benchmarking. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Billion Interactions/s, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM1 EPYC 9684X 2P EPYC 9754 EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 8534P Xeon Platinum 8592+ EPYC 8534PN Xeon Platinum 8380 2P Xeon 6780E 2P Xeon Platinum 8490H Xeon 6766E 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 Xeon 6780E Xeon 6766E 50 100 150 200 250 SE +/- 6.15, N = 12 SE +/- 0.70, N = 9 SE +/- 2.97, N = 13 SE +/- 3.98, N = 15 SE +/- 1.02, N = 8 SE +/- 3.27, N = 13 SE +/- 1.18, N = 7 SE +/- 0.81, N = 7 SE +/- 1.85, N = 12 SE +/- 1.28, N = 12 SE +/- 0.04, N = 6 SE +/- 0.20, N = 6 SE +/- 0.03, N = 6 SE +/- 1.19, N = 15 SE +/- 0.46, N = 15 SE +/- 0.61, N = 5 SE +/- 0.05, N = 5 SE +/- 0.84, N = 15 SE +/- 0.12, N = 4 SE +/- 0.38, N = 9 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 236.61 229.91 217.85 214.62 199.73 151.95 136.95 133.63 114.90 108.40 107.35 106.40 98.21 85.29 84.35 82.85 76.60 70.84 63.62 47.26 43.55 39.17 1. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm Billion Interactions/s Per Watt
OpenBenchmarking.org Billion Interactions/s Per Watt, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM1 EPYC 9754 EPYC 8534PN EPYC 8534P EPYC 9754 2P EPYC 9654 EPYC 9654 2P EPYC 9684X EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P Xeon 6766E Xeon 6780E 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon Platinum 8380 Xeon Platinum 8380 2P 0.3179 0.6358 0.9537 1.2716 1.5895 1.413 1.043 1.008 0.732 0.706 0.655 0.640 0.611 0.503 0.445 0.324 0.291 0.280 0.262 0.255 0.249 0.236 0.231 0.228 0.222 0.206 0.204 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better miniBUDE 20210901 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 EPYC 9654 Xeon 6780E Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H EPYC 9754 2P Xeon 6766E 2P EPYC 9654 2P Xeon 6780E 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 17.11 / Avg: 94.19 / Max: 135.5 Min: 18.42 / Avg: 106.48 / Max: 155.79 Min: 43.57 / Avg: 157.01 / Max: 182.38 Min: 19.01 / Avg: 162.7 / Max: 323.16 Min: 26.83 / Avg: 194.1 / Max: 318.58 Min: 43.42 / Avg: 196.03 / Max: 231.39 Min: 52.99 / Avg: 229.13 / Max: 286.06 Min: 29.7 / Avg: 237.41 / Max: 409.16 Min: 21.92 / Avg: 239.11 / Max: 377.96 Min: 71.42 / Avg: 242.48 / Max: 294.63 Min: 72.35 / Avg: 253.17 / Max: 324.27 Min: 60.74 / Avg: 255.91 / Max: 343.2 Min: 39.64 / Avg: 293.27 / Max: 587.35 Min: 105.45 / Avg: 300.28 / Max: 380.11 Min: 53.82 / Avg: 332.37 / Max: 634.66 Min: 107.92 / Avg: 357 / Max: 473.69 Min: 64.39 / Avg: 387.38 / Max: 798.84 Min: 63.44 / Avg: 396.77 / Max: 754.62 Min: 100.78 / Avg: 417.23 / Max: 578.66 Min: 149.89 / Avg: 459.2 / Max: 669.79 Min: 179.1 / Avg: 476.05 / Max: 713.49 Min: 188.58 / Avg: 496.58 / Max: 737.62
OpenBenchmarking.org GFInst/s, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM1 EPYC 9684X 2P EPYC 9754 EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 8534P Xeon Platinum 8592+ EPYC 8534PN Xeon Platinum 8380 2P Xeon 6780E 2P Xeon Platinum 8490H Xeon 6766E 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 Xeon 6780E Xeon 6766E 1300 2600 3900 5200 6500 SE +/- 153.85, N = 12 SE +/- 17.54, N = 9 SE +/- 74.36, N = 13 SE +/- 99.42, N = 15 SE +/- 25.44, N = 8 SE +/- 81.83, N = 13 SE +/- 29.52, N = 7 SE +/- 20.15, N = 7 SE +/- 46.19, N = 12 SE +/- 32.01, N = 12 SE +/- 0.97, N = 6 SE +/- 5.07, N = 6 SE +/- 0.82, N = 6 SE +/- 29.84, N = 15 SE +/- 11.58, N = 15 SE +/- 15.23, N = 5 SE +/- 1.36, N = 5 SE +/- 21.10, N = 15 SE +/- 2.91, N = 4 SE +/- 9.38, N = 9 SE +/- 0.29, N = 3 SE +/- 0.46, N = 3 5915.25 5747.82 5446.16 5365.42 4993.34 3798.70 3423.63 3340.69 2872.56 2709.93 2683.83 2659.88 2455.16 2132.24 2108.68 2071.19 1915.04 1771.07 1590.50 1181.49 1088.83 979.28 1. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
x265 This is a simple test of the x265 encoder run on the CPU with 1080p and 4K options for H.265 video encode performance with x265. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better x265 3.6 Video Input: Bosphorus 4K Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8490H 2P EPYC 9684X EPYC 9684X 2P EPYC 9654 Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon Max 9468 EPYC 9654 2P EPYC 9754 EPYC 8534P EPYC 9754 2P Xeon Max 9480 2P EPYC 8534PN Xeon Max 9468 2P Xeon 6766E 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P 9 18 27 36 45 SE +/- 0.09, N = 4 SE +/- 0.06, N = 4 SE +/- 0.11, N = 4 SE +/- 0.20, N = 4 SE +/- 0.11, N = 3 SE +/- 0.10, N = 3 SE +/- 0.08, N = 3 SE +/- 0.39, N = 3 SE +/- 0.05, N = 3 SE +/- 0.13, N = 3 SE +/- 0.08, N = 3 SE +/- 0.09, N = 3 SE +/- 0.02, N = 3 SE +/- 0.06, N = 3 SE +/- 0.06, N = 3 SE +/- 0.10, N = 3 SE +/- 0.06, N = 3 SE +/- 0.12, N = 3 SE +/- 0.22, N = 6 SE +/- 0.13, N = 3 SE +/- 0.11, N = 3 SE +/- 0.17, N = 3 39.31 38.93 37.86 37.04 32.12 31.95 31.72 31.40 30.80 29.76 29.05 28.81 28.16 28.13 28.11 27.90 27.74 27.26 22.07 21.30 20.22 19.55 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better x265 3.6 Video Input: Bosphorus 4K EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon 6766E 2P EPYC 9754 2P Xeon Max 9480 Xeon 6780E 2P Xeon Max 9468 Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.0765 0.153 0.2295 0.306 0.3825 0.340 0.318 0.251 0.250 0.242 0.218 0.195 0.178 0.161 0.146 0.138 0.130 0.120 0.115 0.113 0.109 0.108 0.108 0.103 0.090 0.059 0.059 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better x265 3.6 CPU Power Consumption Monitor Xeon 6766E EPYC 8534PN Xeon 6780E EPYC 8534P EPYC 9754 EPYC 9654 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 EPYC 9754 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 90 180 270 360 450 Min: 43.65 / Avg: 80.41 / Max: 103.35 Min: 15.46 / Avg: 81.58 / Max: 101.27 Min: 43.45 / Avg: 88.07 / Max: 115.63 Min: 17.3 / Avg: 88.38 / Max: 113.06 Min: 20.91 / Avg: 112.7 / Max: 141.97 Min: 28.94 / Avg: 145.28 / Max: 189.25 Min: 104.34 / Avg: 160.26 / Max: 184.55 Min: 32.43 / Avg: 164.71 / Max: 212.57 Min: 105.18 / Avg: 169.56 / Max: 195.88 Min: 52.94 / Avg: 203.52 / Max: 247.27 Min: 40.31 / Avg: 216.07 / Max: 262.6 Min: 34.21 / Avg: 219.3 / Max: 292.58 Min: 59.53 / Avg: 234.38 / Max: 293.23 Min: 75.63 / Avg: 256.33 / Max: 294.6 Min: 72.13 / Avg: 261.62 / Max: 312.67 Min: 58.57 / Avg: 267.17 / Max: 327.8 Min: 64.95 / Avg: 296.42 / Max: 367.45 Min: 99.99 / Avg: 298.87 / Max: 362.74 Min: 76.83 / Avg: 359.11 / Max: 458.52 Min: 164.29 / Avg: 409.49 / Max: 491.55 Min: 195.59 / Avg: 458.96 / Max: 515.27 Min: 200.11 / Avg: 471.43 / Max: 532.57
Timed FFmpeg Compilation This test times how long it takes to build the FFmpeg multimedia library. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Timed FFmpeg Compilation 7.0 Time To Compile Xeon Platinum 8592+ 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8490H 2P EPYC 9684X Xeon Max 9480 2P EPYC 9654 Xeon Max 9468 2P EPYC 9754 2P Xeon Platinum 8592+ EPYC 9754 Xeon 6780E 2P Xeon Platinum 8490H Xeon Max 9480 Xeon 6780E Xeon 6766E 2P Xeon Platinum 8380 2P Xeon Max 9468 EPYC 8534P Xeon 6766E EPYC 8534PN Xeon Platinum 8380 8 16 24 32 40 SE +/- 0.05, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.09, N = 3 SE +/- 0.30, N = 3 SE +/- 0.07, N = 3 SE +/- 0.11, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.09, N = 3 SE +/- 0.01, N = 3 SE +/- 0.06, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.06, N = 3 17.39 17.84 18.09 19.51 19.99 20.01 20.49 20.79 21.22 21.36 23.22 23.32 24.06 24.84 25.24 25.71 25.94 26.46 27.54 27.85 28.42 34.29 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Timed FFmpeg Compilation 7.0 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E EPYC 8534P Xeon 6780E EPYC 9754 EPYC 9654 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 17.2 / Avg: 81.55 / Max: 145.14 Min: 43.59 / Avg: 88.1 / Max: 260.1 Min: 17.56 / Avg: 90.17 / Max: 171.06 Min: 43.39 / Avg: 101.96 / Max: 321.98 Min: 21.46 / Avg: 124.42 / Max: 336.98 Min: 28.44 / Avg: 151.49 / Max: 345.46 Min: 103.98 / Avg: 158.14 / Max: 527.88 Min: 32.76 / Avg: 171.98 / Max: 410.22 Min: 55.1 / Avg: 172.28 / Max: 647.64 Min: 52.91 / Avg: 184.1 / Max: 291.84 Min: 34.1 / Avg: 198.27 / Max: 377.58 Min: 57.21 / Avg: 220.08 / Max: 376.48 Min: 39.64 / Avg: 238.4 / Max: 679.1 Min: 74.35 / Avg: 250.86 / Max: 374.99 Min: 81.57 / Avg: 251.37 / Max: 376.37 Min: 55.22 / Avg: 281.94 / Max: 701.16 Min: 98.24 / Avg: 301.65 / Max: 581.66 Min: 65.37 / Avg: 314.5 / Max: 819.41 Min: 85.35 / Avg: 348.01 / Max: 755.27 Min: 157.82 / Avg: 400.67 / Max: 753.15 Min: 198.1 / Avg: 467.2 / Max: 752.99 Min: 208.48 / Avg: 470.97 / Max: 753.46
Pennant Pennant is an application focused on hydrodynamics on general unstructured meshes in 2D. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Hydro Cycle Time - Seconds, Fewer Is Better Pennant 1.0.1 Test: leblancbig EPYC 9684X 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P EPYC 9684X EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 3 6 9 12 15 SE +/- 0.055800, N = 15 SE +/- 0.022246, N = 15 SE +/- 0.013568, N = 6 SE +/- 0.063684, N = 15 SE +/- 0.008354, N = 7 SE +/- 0.019069, N = 15 SE +/- 0.010204, N = 7 SE +/- 0.115724, N = 15 SE +/- 0.115702, N = 15 SE +/- 0.115723, N = 15 SE +/- 0.022698, N = 7 SE +/- 0.037316, N = 9 SE +/- 0.022393, N = 7 SE +/- 0.023904, N = 7 SE +/- 0.025580, N = 6 SE +/- 0.057580, N = 6 SE +/- 0.045713, N = 13 SE +/- 0.010812, N = 4 2.076835 2.084384 2.159459 2.312977 2.527363 2.529807 2.881099 3.360453 3.654303 3.754835 4.182707 4.702557 4.955865 5.114161 5.930957 6.541768 6.852712 11.852130 1. (CXX) g++ options: -fopenmp -lmpi_cxx -lmpi CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Pennant 1.0.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 16.11 / Avg: 95.67 / Max: 145.23 Min: 18.14 / Avg: 107.18 / Max: 164.25 Min: 20.01 / Avg: 156.67 / Max: 328.66 Min: 27.07 / Avg: 170.76 / Max: 330.73 Min: 25.17 / Avg: 191.11 / Max: 377.76 Min: 33.15 / Avg: 196.35 / Max: 409.53 Min: 52.7 / Avg: 215.77 / Max: 289.7 Min: 57.31 / Avg: 222.1 / Max: 377.34 Min: 74.45 / Avg: 234.93 / Max: 375.6 Min: 71.8 / Avg: 241.62 / Max: 375.61 Min: 41.42 / Avg: 291.74 / Max: 665.7 Min: 54.46 / Avg: 322.18 / Max: 670.5 Min: 64.47 / Avg: 342.65 / Max: 753.04 Min: 100.92 / Avg: 344.84 / Max: 581.63 Min: 63.33 / Avg: 370.49 / Max: 818.97 Min: 144.69 / Avg: 414.55 / Max: 753.5 Min: 187.85 / Avg: 448.22 / Max: 752.01 Min: 191.18 / Avg: 450.17 / Max: 753.29
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 1 - Model: AlexNet Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ EPYC 9754 EPYC 8534P EPYC 8534PN Xeon Max 9480 2P EPYC 9654 Xeon 6780E Xeon Platinum 8592+ 2P Xeon 6766E EPYC 9684X EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P 13 26 39 52 65 SE +/- 0.08, N = 8 SE +/- 0.76, N = 15 SE +/- 1.22, N = 15 SE +/- 0.05, N = 7 SE +/- 0.57, N = 15 SE +/- 0.44, N = 15 SE +/- 0.33, N = 15 SE +/- 0.66, N = 15 SE +/- 0.48, N = 15 SE +/- 0.05, N = 7 SE +/- 0.04, N = 7 SE +/- 0.52, N = 15 SE +/- 0.18, N = 7 SE +/- 0.08, N = 7 SE +/- 0.45, N = 15 SE +/- 0.07, N = 6 SE +/- 0.07, N = 7 SE +/- 0.10, N = 6 SE +/- 0.17, N = 15 SE +/- 0.15, N = 9 SE +/- 0.19, N = 5 SE +/- 0.12, N = 15 56.53 48.28 43.67 42.05 37.76 37.57 37.26 35.50 35.08 34.51 34.35 33.21 32.56 31.99 30.76 29.56 26.23 19.92 19.88 18.59 18.34 16.24 images/sec Per Watt
OpenBenchmarking.org images/sec Per Watt, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 1 - Model: AlexNet EPYC 8534PN EPYC 8534P Xeon 6780E Xeon 6766E EPYC 9754 Xeon Platinum 8490H Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9654 Xeon Max 9468 Xeon Max 9480 EPYC 9684X Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon Max 9468 2P Xeon 6766E 2P EPYC 9754 2P Xeon Max 9480 2P EPYC 9654 2P EPYC 9684X 2P 0.1287 0.2574 0.3861 0.5148 0.6435 0.572 0.538 0.407 0.404 0.375 0.374 0.340 0.286 0.275 0.274 0.234 0.181 0.154 0.111 0.106 0.094 0.094 0.087 0.083 0.080 0.071 0.065 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better TensorFlow 2.16.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 120 240 360 480 600 Min: 15.49 / Avg: 60.05 / Max: 115.86 Min: 16.75 / Avg: 64.2 / Max: 124.93 Min: 43.59 / Avg: 73.23 / Max: 123.49 Min: 43.45 / Avg: 78.62 / Max: 138.98 Min: 19.64 / Avg: 93.48 / Max: 194.83 Min: 26.55 / Avg: 118.55 / Max: 230.07 Min: 49.46 / Avg: 123.85 / Max: 241.53 Min: 26.69 / Avg: 124.1 / Max: 239.48 Min: 32.86 / Avg: 144.8 / Max: 258 Min: 59.91 / Avg: 151.32 / Max: 308.57 Min: 75.83 / Avg: 176.45 / Max: 326.47 Min: 73.08 / Avg: 186.7 / Max: 334.25 Min: 104 / Avg: 187.73 / Max: 277.66 Min: 105.47 / Avg: 195.16 / Max: 299.02 Min: 38.63 / Avg: 224.78 / Max: 359.48 Min: 97.99 / Avg: 242.28 / Max: 449.95 Min: 77.54 / Avg: 276.92 / Max: 490.49 Min: 56.56 / Avg: 280.47 / Max: 458.42 Min: 62.19 / Avg: 307.72 / Max: 505.41 Min: 146.77 / Avg: 355.03 / Max: 612.45 Min: 190.72 / Avg: 403 / Max: 643.72 Min: 187.03 / Avg: 414.71 / Max: 652.72
NAS Parallel Benchmarks NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.C EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9754 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 80K 160K 240K 320K 400K SE +/- 1666.53, N = 6 SE +/- 1929.89, N = 5 SE +/- 1528.09, N = 5 SE +/- 1178.78, N = 6 SE +/- 2064.71, N = 6 SE +/- 560.20, N = 4 SE +/- 420.81, N = 4 SE +/- 679.14, N = 4 SE +/- 126.50, N = 4 SE +/- 72.03, N = 4 SE +/- 31.76, N = 4 SE +/- 33.85, N = 3 SE +/- 150.79, N = 3 SE +/- 9.58, N = 3 SE +/- 67.02, N = 3 SE +/- 23.20, N = 3 SE +/- 14.31, N = 3 SE +/- 11.99, N = 3 390520.93 256584.67 228184.68 225311.82 214017.18 147311.14 141954.15 135939.25 127519.37 125191.11 109416.42 90420.25 71245.26 69776.41 58815.36 52745.64 52684.93 37567.61 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.C EPYC 9684X 2P EPYC 9684X EPYC 9654 2P EPYC 9754 EPYC 9754 2P EPYC 9654 Xeon Platinum 8592+ 2P EPYC 8534PN Xeon Platinum 8592+ EPYC 8534P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8380 2P Xeon Max 9468 Xeon Platinum 8380 200 400 600 800 1000 916.62 882.53 649.64 590.51 567.85 547.32 515.52 441.50 411.62 393.05 275.88 251.86 243.32 240.55 229.75 203.59 189.97 151.21 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 17.36 / Avg: 119.33 / Max: 138.69 Min: 18.23 / Avg: 134.2 / Max: 159.66 Min: 22.12 / Avg: 215.95 / Max: 323.68 Min: 26 / Avg: 228.74 / Max: 319.02 Min: 52.53 / Avg: 248.45 / Max: 283.18 Min: 32.04 / Avg: 255.3 / Max: 409.27 Min: 26.78 / Avg: 265.82 / Max: 376.59 Min: 60.29 / Avg: 296.18 / Max: 370.22 Min: 74.98 / Avg: 303.7 / Max: 376.9 Min: 73.26 / Avg: 309.61 / Max: 365.04 Min: 41.44 / Avg: 376.89 / Max: 668.41 Min: 56.6 / Avg: 394.96 / Max: 670.11 Min: 63.25 / Avg: 426.04 / Max: 818.12 Min: 68.95 / Avg: 442.63 / Max: 754.48 Min: 101.91 / Avg: 444.12 / Max: 565.31 Min: 158.17 / Avg: 533.96 / Max: 753.09 Min: 187.62 / Avg: 558.68 / Max: 751.91 Min: 183.64 / Avg: 563.63 / Max: 754.82
PyBench This test profile reports the total time of the different average timed test results from PyBench. PyBench reports average test times for different functions such as BuiltinFunctionCalls and NestedForLoops, with this total result providing a rough estimate as to Python's average performance on a given system. This test profile runs PyBench each time for 20 rounds. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Milliseconds, Fewer Is Better PyBench 2018-02-16 Total For Average Test Times Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Max 9468 Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P EPYC 9684X 2P EPYC 9684X EPYC 9654 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9754 2P Xeon Platinum 8380 EPYC 9754 EPYC 8534PN EPYC 8534P Xeon 6780E 2P Xeon 6780E Xeon 6766E Xeon 6766E 2P 200 400 600 800 1000 SE +/- 0.41, N = 4 SE +/- 0.48, N = 4 SE +/- 0.58, N = 3 SE +/- 0.33, N = 3 SE +/- 0.67, N = 3 SE +/- 0.67, N = 3 SE +/- 0.33, N = 3 SE +/- 1.15, N = 3 SE +/- 1.20, N = 3 SE +/- 2.08, N = 3 SE +/- 2.19, N = 3 SE +/- 1.86, N = 3 SE +/- 0.33, N = 3 SE +/- 1.33, N = 3 SE +/- 2.85, N = 3 SE +/- 0.58, N = 3 SE +/- 2.40, N = 3 SE +/- 1.45, N = 3 SE +/- 0.58, N = 3 SE +/- 0.88, N = 3 SE +/- 1.15, N = 3 SE +/- 0.33, N = 3 691 692 762 763 763 763 765 767 818 818 818 822 971 972 973 974 974 976 1005 1007 1117 1119 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better PyBench 2018-02-16 CPU Power Consumption Monitor Xeon 6766E Xeon 6780E EPYC 8534PN EPYC 8534P EPYC 9754 Xeon Platinum 8592+ EPYC 9654 Xeon Platinum 8490H Xeon Platinum 8380 EPYC 9684X Xeon 6766E 2P Xeon 6780E 2P Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 60 120 180 240 300 Min: 43.58 / Avg: 43.95 / Max: 46.08 Min: 43.46 / Avg: 45.25 / Max: 46 Min: 14.86 / Avg: 47.57 / Max: 53.35 Min: 15.93 / Avg: 48.41 / Max: 54.07 Min: 19.44 / Avg: 72.53 / Max: 81.29 Min: 33.53 / Avg: 85.96 / Max: 134.17 Min: 28.15 / Avg: 88.41 / Max: 100.93 Min: 58.93 / Avg: 88.45 / Max: 123.25 Min: 49.36 / Avg: 92.52 / Max: 101.33 Min: 33.11 / Avg: 99.97 / Max: 113.9 Min: 103.97 / Avg: 107.3 / Max: 109.09 Min: 106.05 / Avg: 107.95 / Max: 109.82 Min: 74.61 / Avg: 114.52 / Max: 125.95 Min: 83.03 / Avg: 121.48 / Max: 133.99 Min: 38.38 / Avg: 175.13 / Max: 196.99 Min: 98.37 / Avg: 177.98 / Max: 196.22 Min: 79.78 / Avg: 194.14 / Max: 289.01 Min: 57.47 / Avg: 207.73 / Max: 239.74 Min: 163.51 / Avg: 228.16 / Max: 305.07 Min: 65.07 / Avg: 232.2 / Max: 264.86 Min: 206.86 / Avg: 276.72 / Max: 343.62 Min: 205.54 / Avg: 276.95 / Max: 327.1
NAS Parallel Benchmarks NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: LU.C EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9754 EPYC 9654 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H EPYC 8534P Xeon Max 9480 EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 130K 260K 390K 520K 650K SE +/- 5800.63, N = 15 SE +/- 1879.23, N = 6 SE +/- 4162.49, N = 6 SE +/- 726.00, N = 5 SE +/- 1849.65, N = 6 SE +/- 2566.89, N = 15 SE +/- 789.76, N = 5 SE +/- 398.40, N = 5 SE +/- 411.20, N = 5 SE +/- 441.66, N = 5 SE +/- 213.44, N = 5 SE +/- 155.95, N = 4 SE +/- 130.20, N = 4 SE +/- 314.42, N = 4 SE +/- 37.04, N = 4 SE +/- 274.93, N = 3 SE +/- 71.37, N = 3 SE +/- 123.65, N = 3 588329.32 551955.55 472931.04 455312.09 362511.43 279397.45 269047.28 261127.09 257384.55 246822.52 232012.77 185954.54 135125.15 133284.46 131606.86 130839.12 126171.52 98557.16 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: LU.C EPYC 9754 2P EPYC 9684X EPYC 9754 EPYC 9684X 2P EPYC 9654 2P EPYC 9654 EPYC 8534PN Xeon Platinum 8592+ 2P EPYC 8534P Xeon Platinum 8592+ Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Max 9480 2P Xeon Max 9468 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P Xeon Platinum 8380 400 800 1200 1600 2000 1808.53 1613.95 1409.22 1398.79 1335.18 1333.71 1212.35 1127.30 1120.43 960.89 522.54 503.44 478.19 459.15 449.46 449.20 445.10 422.75 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 17.84 / Avg: 107.92 / Max: 134.69 Min: 17.95 / Avg: 118.96 / Max: 149.72 Min: 21.38 / Avg: 198.26 / Max: 322.68 Min: 26.14 / Avg: 201.73 / Max: 306.76 Min: 32.65 / Avg: 224.61 / Max: 370.12 Min: 51.86 / Avg: 233.13 / Max: 282.84 Min: 26.25 / Avg: 241.46 / Max: 364.07 Min: 59.44 / Avg: 268.4 / Max: 328.81 Min: 74.49 / Avg: 280.72 / Max: 336.16 Min: 76.87 / Avg: 292.98 / Max: 353.23 Min: 40.45 / Avg: 325.31 / Max: 668.91 Min: 56.83 / Avg: 354.21 / Max: 638.12 Min: 65.04 / Avg: 394.59 / Max: 765.59 Min: 67.67 / Avg: 403.9 / Max: 715.05 Min: 99.57 / Avg: 417.78 / Max: 563.12 Min: 157.28 / Avg: 499.73 / Max: 701.36 Min: 194.99 / Avg: 537.56 / Max: 710.4 Min: 196.19 / Avg: 538.25 / Max: 736.91
ASTC Encoder ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 4.7 Preset: Thorough EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9754 Xeon Max 9480 2P Xeon 6780E 2P EPYC 9684X Xeon Max 9468 2P Xeon 6766E 2P EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6780E Xeon Max 9480 EPYC 8534P Xeon 6766E EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8380 30 60 90 120 150 SE +/- 1.20, N = 5 SE +/- 1.46, N = 12 SE +/- 0.73, N = 15 SE +/- 0.55, N = 5 SE +/- 0.19, N = 5 SE +/- 0.01, N = 5 SE +/- 0.03, N = 5 SE +/- 0.85, N = 4 SE +/- 0.06, N = 5 SE +/- 0.05, N = 4 SE +/- 0.42, N = 4 SE +/- 0.01, N = 4 SE +/- 0.05, N = 4 SE +/- 0.01, N = 4 SE +/- 0.01, N = 3 SE +/- 0.03, N = 4 SE +/- 0.02, N = 4 SE +/- 0.00, N = 3 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 121.30 116.12 105.41 97.94 86.98 86.01 83.34 79.29 73.38 72.33 71.02 65.30 54.45 51.30 45.21 42.98 41.02 40.69 37.96 37.01 30.03 15.50 1. (CXX) g++ options: -O3 -flto -pthread MT/s Per Watt
OpenBenchmarking.org MT/s Per Watt, More Is Better ASTC Encoder 4.7 Preset: Thorough EPYC 9754 EPYC 9754 2P EPYC 8534PN EPYC 8534P EPYC 9654 EPYC 9684X EPYC 9654 2P EPYC 9684X 2P Xeon 6766E 2P Xeon 6780E 2P Xeon 6766E Xeon Platinum 8592+ 2P Xeon 6780E Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9480 Xeon Max 9468 2P Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8380 0.1136 0.2272 0.3408 0.4544 0.568 0.505 0.456 0.384 0.359 0.338 0.330 0.319 0.316 0.310 0.299 0.290 0.262 0.259 0.244 0.214 0.203 0.179 0.164 0.153 0.139 0.074 0.068 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ASTC Encoder 4.7 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8380 Xeon 6766E 2P Xeon Platinum 8490H Xeon Max 9480 Xeon 6780E 2P Xeon Max 9468 EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 15.09 / Avg: 98.95 / Max: 145.65 Min: 15.96 / Avg: 114.12 / Max: 173.87 Min: 43.64 / Avg: 140.53 / Max: 230.96 Min: 20.4 / Avg: 170.2 / Max: 353.81 Min: 43.44 / Avg: 174.24 / Max: 297.52 Min: 28.71 / Avg: 193.35 / Max: 335.29 Min: 33.53 / Avg: 222.24 / Max: 407.18 Min: 34.05 / Avg: 223.18 / Max: 369.47 Min: 49.99 / Avg: 228.75 / Max: 289.57 Min: 103.99 / Avg: 228.94 / Max: 433.24 Min: 60.85 / Avg: 239.45 / Max: 374.18 Min: 78.02 / Avg: 261.71 / Max: 374.53 Min: 106.82 / Avg: 264.73 / Max: 530.48 Min: 72.98 / Avg: 265.38 / Max: 375.06 Min: 38.72 / Avg: 265.99 / Max: 577.68 Min: 55.87 / Avg: 330.17 / Max: 623.66 Min: 63.05 / Avg: 367.67 / Max: 734.43 Min: 76.68 / Avg: 373.98 / Max: 732.13 Min: 99.18 / Avg: 403.9 / Max: 572.69 Min: 147.85 / Avg: 428.35 / Max: 731.41 Min: 195.71 / Avg: 466.49 / Max: 741.06 Min: 204.94 / Avg: 472.75 / Max: 744.21
NAS Parallel Benchmarks NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: IS.D EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9754 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H EPYC 8534P EPYC 8534PN Xeon Platinum 8380 3K 6K 9K 12K 15K SE +/- 119.22, N = 6 SE +/- 76.70, N = 10 SE +/- 85.04, N = 8 SE +/- 155.00, N = 15 SE +/- 29.95, N = 5 SE +/- 35.67, N = 5 SE +/- 17.53, N = 5 SE +/- 8.92, N = 4 SE +/- 23.92, N = 4 SE +/- 4.96, N = 4 SE +/- 12.99, N = 4 SE +/- 8.35, N = 4 SE +/- 18.08, N = 4 SE +/- 15.55, N = 4 SE +/- 6.56, N = 3 SE +/- 9.85, N = 4 SE +/- 14.39, N = 4 SE +/- 11.05, N = 3 12296.27 10139.69 9771.41 7437.81 6346.15 5702.13 5229.41 4059.70 3960.28 3819.83 3816.70 3345.45 3076.66 3050.16 2865.72 2746.20 2718.94 2266.32 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: IS.D EPYC 9754 EPYC 9684X 2P EPYC 9654 2P EPYC 9684X EPYC 9754 2P EPYC 9654 EPYC 8534PN EPYC 8534P Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9480 Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 8 16 24 32 40 36.454 35.302 32.063 32.045 31.780 29.667 27.818 25.260 18.332 18.030 11.909 11.814 11.707 10.738 8.646 8.360 7.333 7.258 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 9.26 / Avg: 97.74 / Max: 132.42 Min: 19.1 / Avg: 108.72 / Max: 150.47 Min: 20.68 / Avg: 156.42 / Max: 284.24 Min: 25.79 / Avg: 176.27 / Max: 304.26 Min: 32.82 / Avg: 198.04 / Max: 362.78 Min: 54.69 / Avg: 211.05 / Max: 280.75 Min: 25.68 / Avg: 225.16 / Max: 371.3 Min: 60.58 / Avg: 242.56 / Max: 344.06 Min: 75.18 / Avg: 256.12 / Max: 347.73 Min: 75.39 / Avg: 262.81 / Max: 362.43 Min: 41.36 / Avg: 307.47 / Max: 647.96 Min: 54.94 / Avg: 316.24 / Max: 616.93 Min: 66.48 / Avg: 348.31 / Max: 715.01 Min: 103.51 / Avg: 386.95 / Max: 547.42 Min: 67.06 / Avg: 405.72 / Max: 726.27 Min: 145.91 / Avg: 473.71 / Max: 694.3 Min: 196.72 / Avg: 520.51 / Max: 703.63 Min: 196.74 / Avg: 526.3 / Max: 723.61
Timed Mesa Compilation This test profile times how long it takes to compile Mesa with Meson/Ninja. For minimizing build dependencies and avoid versioning conflicts, test this is just the core Mesa build without LLVM or the extra Gallium3D/Mesa drivers enabled. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Timed Mesa Compilation 24.0 Time To Compile Xeon Platinum 8592+ 2P EPYC 9684X 2P EPYC 9654 2P EPYC 9684X EPYC 9654 Xeon Platinum 8592+ Xeon Max 9480 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9480 EPYC 9754 Xeon Max 9468 EPYC 8534P Xeon Platinum 8380 2P EPYC 8534PN Xeon 6780E Xeon 6780E 2P Xeon Platinum 8380 Xeon 6766E 2P Xeon 6766E 6 12 18 24 30 SE +/- 0.05, N = 4 SE +/- 0.08, N = 4 SE +/- 0.02, N = 4 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.07, N = 3 SE +/- 0.17, N = 3 SE +/- 0.12, N = 3 SE +/- 0.07, N = 3 SE +/- 0.07, N = 3 SE +/- 0.12, N = 3 SE +/- 0.08, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.05, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 SE +/- 0.06, N = 3 SE +/- 0.05, N = 3 SE +/- 0.08, N = 3 SE +/- 0.06, N = 3 16.18 16.40 16.41 16.67 16.96 17.26 18.14 18.29 18.47 19.23 19.56 19.65 19.91 20.18 21.04 21.44 21.53 23.41 23.46 23.66 25.88 25.89 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Timed Mesa Compilation 24.0 CPU Power Consumption Monitor Xeon 6766E EPYC 8534PN EPYC 8534P Xeon 6780E EPYC 9754 EPYC 9654 Xeon 6766E 2P Xeon 6780E 2P EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8490H Xeon Platinum 8592+ EPYC 9754 2P Xeon Max 9480 Xeon Max 9468 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 43.62 / Avg: 71.87 / Max: 255.49 Min: 16.52 / Avg: 73.01 / Max: 143.27 Min: 18.48 / Avg: 79 / Max: 166.34 Min: 43.31 / Avg: 79.63 / Max: 313.34 Min: 20.4 / Avg: 109.13 / Max: 324.24 Min: 27.72 / Avg: 132.61 / Max: 328.77 Min: 104.21 / Avg: 137.28 / Max: 498.1 Min: 106.87 / Avg: 146.41 / Max: 572.5 Min: 33.71 / Avg: 153.28 / Max: 411.05 Min: 52.69 / Avg: 166.64 / Max: 291.22 Min: 57 / Avg: 172.35 / Max: 378.42 Min: 32.86 / Avg: 176.64 / Max: 377.39 Min: 41.25 / Avg: 215.57 / Max: 638.97 Min: 73.22 / Avg: 228.31 / Max: 375.9 Min: 74.89 / Avg: 232.74 / Max: 375.37 Min: 53.33 / Avg: 251.26 / Max: 671.83 Min: 99.18 / Avg: 266.43 / Max: 584.3 Min: 65.81 / Avg: 287.81 / Max: 802.37 Min: 82.26 / Avg: 324.23 / Max: 753.39 Min: 171.78 / Avg: 374.89 / Max: 753.6 Min: 195.35 / Avg: 442.47 / Max: 753.73 Min: 199.11 / Avg: 446.38 / Max: 751.99
Xcompact3d Incompact3d Xcompact3d Incompact3d is a Fortran-MPI based, finite difference high-performance code for solving the incompressible Navier-Stokes equation and as many as you need scalar transport equations. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 Input: input.i3d 193 Cells Per Direction EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9684X Xeon Max 9468 2P EPYC 9654 EPYC 9754 Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Max 9480 Xeon Platinum 8490H Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 6 12 18 24 30 SE +/- 0.00891960, N = 5 SE +/- 0.04312156, N = 5 SE +/- 0.01633300, N = 4 SE +/- 0.03329814, N = 5 SE +/- 0.04312801, N = 5 SE +/- 0.05695382, N = 15 SE +/- 0.05847189, N = 6 SE +/- 0.01741826, N = 5 SE +/- 0.06577912, N = 5 SE +/- 0.03951909, N = 5 SE +/- 0.02923582, N = 4 SE +/- 0.02691998, N = 4 SE +/- 0.01708632, N = 4 SE +/- 0.01004100, N = 4 SE +/- 0.04445674, N = 4 SE +/- 0.01040414, N = 3 SE +/- 0.10237421, N = 3 SE +/- 0.01216869, N = 3 3.75319901 4.26078596 4.51061893 5.39125967 6.39169817 6.66387510 6.88121279 7.29349117 8.25383511 8.87225914 10.19670660 11.24980160 12.79975820 12.92934130 13.96923760 20.93791390 21.17803070 23.97082840 1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 17.28 / Avg: 114.52 / Max: 138.44 Min: 19.02 / Avg: 124.9 / Max: 149.76 Min: 22.21 / Avg: 200.17 / Max: 312.99 Min: 30.56 / Avg: 216.15 / Max: 325.7 Min: 53.04 / Avg: 237.24 / Max: 284.02 Min: 24.14 / Avg: 251.56 / Max: 369.25 Min: 34.54 / Avg: 254.46 / Max: 403.7 Min: 58.88 / Avg: 278.29 / Max: 373.12 Min: 77.21 / Avg: 283.32 / Max: 372.73 Min: 75.94 / Avg: 288.56 / Max: 366.2 Min: 40.05 / Avg: 313.08 / Max: 632.39 Min: 57.92 / Avg: 371 / Max: 659.72 Min: 100.46 / Avg: 415.93 / Max: 564.56 Min: 78.61 / Avg: 426.75 / Max: 741.39 Min: 64.22 / Avg: 431.35 / Max: 814.6 Min: 151.7 / Avg: 495.02 / Max: 746.22 Min: 187.55 / Avg: 524.04 / Max: 746.57 Min: 196.28 / Avg: 524.59 / Max: 749.74
Y-Cruncher Y-Cruncher is a multi-threaded Pi benchmark capable of computing Pi to trillions of digits. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Y-Cruncher 0.8.3 Pi Digits To Calculate: 1B Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 9654 EPYC 9754 Xeon Platinum 8490H Xeon 6780E 2P EPYC 9684X Xeon Platinum 8380 2P EPYC 9754 2P Xeon Max 9480 Xeon 6766E 2P EPYC 9654 2P Xeon Max 9468 EPYC 9684X 2P EPYC 8534P Xeon 6780E EPYC 8534PN Xeon Platinum 8380 Xeon 6766E 3 6 9 12 15 SE +/- 0.007, N = 6 SE +/- 0.010, N = 6 SE +/- 0.142, N = 12 SE +/- 0.122, N = 15 SE +/- 0.030, N = 5 SE +/- 0.008, N = 5 SE +/- 0.007, N = 4 SE +/- 0.005, N = 5 SE +/- 0.002, N = 4 SE +/- 0.008, N = 4 SE +/- 0.010, N = 5 SE +/- 0.049, N = 4 SE +/- 0.026, N = 5 SE +/- 0.007, N = 4 SE +/- 0.012, N = 4 SE +/- 0.045, N = 5 SE +/- 0.012, N = 4 SE +/- 0.005, N = 4 SE +/- 0.020, N = 4 SE +/- 0.001, N = 4 SE +/- 0.011, N = 4 SE +/- 0.018, N = 4 5.008 5.110 6.266 6.334 6.701 6.743 6.928 7.047 7.468 7.471 7.744 7.915 7.959 7.967 8.196 8.300 8.992 10.205 10.386 10.427 10.712 11.198 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Y-Cruncher 0.8.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 16.4 / Avg: 94.34 / Max: 147.45 Min: 16.07 / Avg: 105.12 / Max: 171.94 Min: 43.6 / Avg: 141.11 / Max: 252.27 Min: 21.57 / Avg: 148.59 / Max: 323.95 Min: 43.37 / Avg: 168.68 / Max: 308.71 Min: 28.49 / Avg: 178.56 / Max: 342.72 Min: 53.53 / Avg: 195.15 / Max: 284.61 Min: 32.02 / Avg: 198.74 / Max: 364.57 Min: 32.32 / Avg: 212.59 / Max: 399.38 Min: 30.38 / Avg: 222.94 / Max: 366.15 Min: 73.1 / Avg: 254.54 / Max: 371.3 Min: 76.02 / Avg: 255.33 / Max: 373.96 Min: 104.12 / Avg: 272.77 / Max: 508.32 Min: 40.66 / Avg: 284.7 / Max: 529.69 Min: 105.25 / Avg: 293.09 / Max: 598.55 Min: 58.9 / Avg: 351.79 / Max: 591.63 Min: 98.98 / Avg: 352.71 / Max: 555.83 Min: 76.49 / Avg: 362.82 / Max: 745.4 Min: 66.34 / Avg: 404.17 / Max: 643.66 Min: 166.8 / Avg: 424.33 / Max: 738.75 Min: 193.6 / Avg: 493.66 / Max: 748.88 Min: 191.23 / Avg: 496.54 / Max: 748.22
Timed ImageMagick Compilation This test times how long it takes to build ImageMagick. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Timed ImageMagick Compilation 6.9.0 Time To Compile Xeon Platinum 8592+ 2P EPYC 9684X 2P EPYC 9684X Xeon Platinum 8592+ EPYC 9654 2P Xeon Platinum 8490H 2P EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H Xeon Max 9480 EPYC 9754 2P Xeon Max 9468 Xeon Platinum 8380 2P EPYC 9754 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 Xeon 6780E 2P Xeon 6780E Xeon 6766E 2P Xeon 6766E 5 10 15 20 25 SE +/- 0.03, N = 4 SE +/- 0.04, N = 4 SE +/- 0.06, N = 4 SE +/- 0.10, N = 4 SE +/- 0.04, N = 4 SE +/- 0.13, N = 4 SE +/- 0.09, N = 4 SE +/- 0.07, N = 4 SE +/- 0.11, N = 4 SE +/- 0.10, N = 4 SE +/- 0.06, N = 4 SE +/- 0.02, N = 4 SE +/- 0.15, N = 4 SE +/- 0.13, N = 4 SE +/- 0.08, N = 4 SE +/- 0.04, N = 3 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 SE +/- 0.11, N = 3 SE +/- 0.05, N = 3 SE +/- 0.08, N = 3 SE +/- 0.08, N = 3 11.49 11.72 12.20 12.60 12.75 13.26 13.47 14.09 14.40 14.72 15.38 15.81 16.08 16.33 16.39 17.92 18.18 18.84 18.85 19.20 20.59 20.91 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Timed ImageMagick Compilation 6.9.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon 6766E 2P EPYC 9684X Xeon 6780E 2P Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9480 Xeon Max 9468 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 16.43 / Avg: 65.12 / Max: 142.84 Min: 17.83 / Avg: 69.48 / Max: 164.08 Min: 43.53 / Avg: 69.56 / Max: 253.22 Min: 43.41 / Avg: 74.98 / Max: 300.96 Min: 21.44 / Avg: 97.43 / Max: 329.22 Min: 28.39 / Avg: 120.39 / Max: 332.54 Min: 103.94 / Avg: 137.3 / Max: 386.07 Min: 32.48 / Avg: 139.23 / Max: 409.7 Min: 107.23 / Avg: 143.22 / Max: 459.28 Min: 51.29 / Avg: 152.46 / Max: 291.86 Min: 26.01 / Avg: 162.46 / Max: 377 Min: 60.5 / Avg: 186.22 / Max: 378.44 Min: 40.62 / Avg: 204.88 / Max: 561.52 Min: 75.62 / Avg: 220.24 / Max: 379.25 Min: 73.52 / Avg: 220.79 / Max: 376.49 Min: 57.57 / Avg: 245.77 / Max: 578.82 Min: 97.82 / Avg: 252.21 / Max: 581.41 Min: 62.42 / Avg: 273.56 / Max: 669.18 Min: 81.81 / Avg: 295.33 / Max: 748.04 Min: 164.76 / Avg: 362.03 / Max: 752.14 Min: 198.51 / Avg: 429.23 / Max: 750.07 Min: 203.42 / Avg: 435.59 / Max: 750.99
m-queens A solver for the N-queens problem with multi-threading support via the OpenMP library. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better m-queens 1.2 Time To Solve EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9754 EPYC 9684X Xeon Platinum 8592+ 2P EPYC 9654 Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon 6780E Xeon Max 9468 2P Xeon 6766E Xeon Platinum 8380 2P EPYC 8534P Xeon Platinum 8592+ EPYC 8534PN Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 6 12 18 24 30 SE +/- 0.031, N = 8 SE +/- 0.064, N = 15 SE +/- 0.093, N = 15 SE +/- 0.034, N = 7 SE +/- 0.035, N = 7 SE +/- 0.021, N = 6 SE +/- 0.025, N = 6 SE +/- 0.028, N = 6 SE +/- 0.019, N = 6 SE +/- 0.021, N = 6 SE +/- 0.028, N = 5 SE +/- 0.032, N = 5 SE +/- 0.021, N = 5 SE +/- 0.039, N = 5 SE +/- 0.066, N = 4 SE +/- 0.019, N = 4 SE +/- 0.030, N = 4 SE +/- 0.029, N = 4 SE +/- 0.025, N = 4 SE +/- 0.035, N = 3 SE +/- 0.031, N = 3 SE +/- 0.023, N = 3 3.981 4.133 4.609 5.512 6.045 6.350 6.950 7.402 7.738 7.881 9.285 10.301 10.840 11.479 12.390 13.250 14.257 14.493 15.211 18.102 21.030 24.175 1. (CXX) g++ options: -fopenmp -O2 -march=native CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better m-queens 1.2 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 EPYC 9654 Xeon 6780E Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8490H EPYC 9684X Xeon Platinum 8592+ EPYC 9754 2P Xeon 6766E 2P EPYC 9654 2P Xeon 6780E 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 15.51 / Avg: 106.99 / Max: 136.77 Min: 19.39 / Avg: 122.5 / Max: 158.25 Min: 43.69 / Avg: 180.19 / Max: 242.1 Min: 20.79 / Avg: 202.95 / Max: 326.11 Min: 28.35 / Avg: 221.83 / Max: 330.6 Min: 43.54 / Avg: 226.9 / Max: 315.05 Min: 54.03 / Avg: 238.8 / Max: 290.84 Min: 75.51 / Avg: 241.11 / Max: 276.68 Min: 77.06 / Avg: 260.39 / Max: 302.34 Min: 62.12 / Avg: 266.4 / Max: 324.62 Min: 31.01 / Avg: 267.9 / Max: 411.07 Min: 32.66 / Avg: 278.57 / Max: 374.34 Min: 41.4 / Avg: 316.2 / Max: 657.62 Min: 105.96 / Avg: 317.88 / Max: 500.32 Min: 52.18 / Avg: 366.51 / Max: 668.54 Min: 108.43 / Avg: 381.31 / Max: 634.03 Min: 63.38 / Avg: 412.8 / Max: 820.43 Min: 99.79 / Avg: 431.33 / Max: 582.73 Min: 201.32 / Avg: 465.55 / Max: 574.13 Min: 77.88 / Avg: 481.61 / Max: 753.74 Min: 165.88 / Avg: 497.84 / Max: 742.37 Min: 180.86 / Avg: 502.47 / Max: 745.73
Intel Open Image Denoise Open Image Denoise is a denoising library for ray-tracing and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 2.2 Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-Only Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9754 2P EPYC 9654 2P Xeon Max 9468 2P EPYC 9684X 2P Xeon Platinum 8592+ Xeon 6780E 2P EPYC 9754 Xeon Platinum 8380 2P Xeon 6766E 2P EPYC 9654 Xeon Platinum 8490H EPYC 9684X Xeon Max 9480 Xeon Max 9468 Xeon 6780E Xeon 6766E Xeon Platinum 8380 EPYC 8534P EPYC 8534PN 1.1453 2.2906 3.4359 4.5812 5.7265 SE +/- 0.03, N = 6 SE +/- 0.01, N = 6 SE +/- 0.02, N = 5 SE +/- 0.01, N = 5 SE +/- 0.01, N = 5 SE +/- 0.02, N = 5 SE +/- 0.00, N = 5 SE +/- 0.00, N = 5 SE +/- 0.00, N = 5 SE +/- 0.03, N = 6 SE +/- 0.00, N = 5 SE +/- 0.01, N = 4 SE +/- 0.01, N = 4 SE +/- 0.00, N = 4 SE +/- 0.02, N = 4 SE +/- 0.00, N = 4 SE +/- 0.00, N = 4 SE +/- 0.00, N = 4 SE +/- 0.00, N = 4 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 5.09 4.48 3.66 3.57 3.49 3.49 3.47 3.32 3.31 3.16 3.10 2.97 2.73 2.71 2.66 2.39 2.27 2.11 1.93 1.81 1.80 1.69 Images / Sec Per Watt
OpenBenchmarking.org Images / Sec Per Watt, More Is Better Intel Open Image Denoise 2.2 Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-Only EPYC 8534PN EPYC 8534P EPYC 9754 Xeon Platinum 8592+ EPYC 9654 Xeon Platinum 8592+ 2P Xeon 6766E EPYC 9754 2P Xeon Platinum 8490H Xeon 6780E EPYC 9684X Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9654 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Max 9480 Xeon Max 9468 EPYC 9684X 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.0036 0.0072 0.0108 0.0144 0.018 0.016 0.015 0.015 0.013 0.012 0.012 0.010 0.010 0.010 0.009 0.009 0.009 0.008 0.008 0.008 0.008 0.008 0.008 0.008 0.007 0.007 0.006 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Intel Open Image Denoise 2.2 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon 6780E Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9684X Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon 6780E 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 17.42 / Avg: 107.91 / Max: 132.12 Min: 18.74 / Avg: 123.7 / Max: 154.19 Min: 43.72 / Avg: 199.4 / Max: 250.2 Min: 21.62 / Avg: 205.54 / Max: 298.26 Min: 52.95 / Avg: 224.45 / Max: 283.1 Min: 27.41 / Avg: 233.36 / Max: 317.91 Min: 43.54 / Avg: 247.22 / Max: 322.66 Min: 26.77 / Avg: 249.97 / Max: 378.66 Min: 59.91 / Avg: 268.44 / Max: 367.04 Min: 32.55 / Avg: 282.53 / Max: 386.33 Min: 76.35 / Avg: 284.74 / Max: 374.26 Min: 73.66 / Avg: 286.68 / Max: 372.63 Min: 105.15 / Avg: 359.38 / Max: 527.77 Min: 40.05 / Avg: 372.52 / Max: 545.78 Min: 100.45 / Avg: 406.6 / Max: 560.45 Min: 57.66 / Avg: 424.59 / Max: 613.78 Min: 108.16 / Avg: 427.37 / Max: 638.82 Min: 80.74 / Avg: 440.63 / Max: 738.2 Min: 165.7 / Avg: 496.66 / Max: 745.69 Min: 63.03 / Avg: 499.92 / Max: 728.7 Min: 193.7 / Avg: 538.98 / Max: 749.89 Min: 208.03 / Avg: 550.26 / Max: 748.26
Result
OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 2.2 Run: RT.hdr_alb_nrm.3840x2160 - Device: CPU-Only Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9754 2P Xeon Max 9468 2P EPYC 9654 2P EPYC 9684X 2P Xeon 6780E 2P Xeon Platinum 8592+ EPYC 9754 Xeon Platinum 8380 2P Xeon 6766E 2P EPYC 9654 Xeon Platinum 8490H EPYC 9684X Xeon Max 9480 Xeon Max 9468 Xeon 6780E Xeon 6766E Xeon Platinum 8380 EPYC 8534P EPYC 8534PN 1.1318 2.2636 3.3954 4.5272 5.659 SE +/- 0.03, N = 6 SE +/- 0.03, N = 6 SE +/- 0.04, N = 5 SE +/- 0.01, N = 5 SE +/- 0.01, N = 5 SE +/- 0.01, N = 5 SE +/- 0.00, N = 5 SE +/- 0.03, N = 5 SE +/- 0.00, N = 5 SE +/- 0.02, N = 5 SE +/- 0.01, N = 5 SE +/- 0.00, N = 4 SE +/- 0.01, N = 4 SE +/- 0.00, N = 4 SE +/- 0.00, N = 4 SE +/- 0.00, N = 4 SE +/- 0.00, N = 4 SE +/- 0.00, N = 4 SE +/- 0.00, N = 4 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 5.03 4.41 3.65 3.59 3.51 3.48 3.45 3.31 3.31 3.12 3.08 2.98 2.73 2.71 2.64 2.38 2.26 2.11 1.93 1.81 1.79 1.69 Images / Sec Per Watt
OpenBenchmarking.org Images / Sec Per Watt, More Is Better Intel Open Image Denoise 2.2 Run: RT.hdr_alb_nrm.3840x2160 - Device: CPU-Only EPYC 8534PN EPYC 9754 EPYC 8534P Xeon Platinum 8592+ EPYC 9654 Xeon Platinum 8592+ 2P Xeon 6766E EPYC 9754 2P Xeon Platinum 8490H Xeon 6780E EPYC 9684X Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9654 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon Max 9480 Xeon Max 9468 EPYC 9684X 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.0036 0.0072 0.0108 0.0144 0.018 0.016 0.015 0.014 0.013 0.012 0.011 0.010 0.010 0.010 0.009 0.009 0.009 0.008 0.008 0.008 0.008 0.008 0.008 0.008 0.007 0.007 0.006 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Intel Open Image Denoise 2.2 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon 6780E Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 EPYC 9684X Xeon Max 9468 Xeon 6766E 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 17.3 / Avg: 107.16 / Max: 132.19 Min: 19.01 / Avg: 124.67 / Max: 154.19 Min: 43.68 / Avg: 198.71 / Max: 249.99 Min: 21.86 / Avg: 203.92 / Max: 294.28 Min: 53.39 / Avg: 224.83 / Max: 282.65 Min: 25.93 / Avg: 230.11 / Max: 317.81 Min: 43.49 / Avg: 246.81 / Max: 322.99 Min: 26.29 / Avg: 249.88 / Max: 378.95 Min: 60.35 / Avg: 271.32 / Max: 366.61 Min: 76.26 / Avg: 283.96 / Max: 374.08 Min: 32.59 / Avg: 284.36 / Max: 386.47 Min: 73.14 / Avg: 287.37 / Max: 372.13 Min: 104.93 / Avg: 363.85 / Max: 529.39 Min: 40.47 / Avg: 369.54 / Max: 546.36 Min: 100.7 / Avg: 406.71 / Max: 559.15 Min: 107.69 / Avg: 422.71 / Max: 638.83 Min: 56.38 / Avg: 425.53 / Max: 615.08 Min: 72.33 / Avg: 443.88 / Max: 736.09 Min: 177.66 / Avg: 501.77 / Max: 743.78 Min: 66.11 / Avg: 502.25 / Max: 723.91 Min: 193.14 / Avg: 539.15 / Max: 749.93 Min: 203.58 / Avg: 542.52 / Max: 747.25
NAS Parallel Benchmarks NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: CG.C Xeon Platinum 8592+ 2P EPYC 9684X 2P EPYC 9654 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ EPYC 9754 2P EPYC 9684X Xeon Platinum 8490H 2P EPYC 9654 EPYC 9754 Xeon Platinum 8380 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H Xeon Platinum 8380 EPYC 8534P EPYC 8534PN 20K 40K 60K 80K 100K SE +/- 271.54, N = 6 SE +/- 661.05, N = 7 SE +/- 779.47, N = 7 SE +/- 560.41, N = 15 SE +/- 545.92, N = 15 SE +/- 328.18, N = 8 SE +/- 545.82, N = 5 SE +/- 386.70, N = 9 SE +/- 178.23, N = 8 SE +/- 393.67, N = 15 SE +/- 480.71, N = 15 SE +/- 28.62, N = 7 SE +/- 260.16, N = 15 SE +/- 313.88, N = 7 SE +/- 243.41, N = 12 SE +/- 34.34, N = 6 SE +/- 126.79, N = 6 SE +/- 207.19, N = 6 98764.93 88487.91 84477.72 66492.44 65396.22 63760.45 63265.01 62071.10 58128.24 53182.77 43799.78 38906.13 35234.06 35116.55 33811.63 22655.32 22229.94 21684.80 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: CG.C Xeon Platinum 8592+ EPYC 9684X EPYC 9654 Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9684X 2P EPYC 9754 EPYC 9754 2P EPYC 8534PN EPYC 8534P Xeon Max 9468 Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P Xeon Platinum 8380 Xeon Platinum 8380 2P 90 180 270 360 450 401.38 384.09 356.27 305.56 302.72 295.11 283.84 254.62 244.61 226.89 157.12 156.61 155.78 154.97 153.93 145.15 119.97 119.68 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9654 EPYC 9754 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 16.9 / Avg: 88.65 / Max: 135.53 Min: 17.31 / Avg: 97.98 / Max: 151 Min: 26.59 / Avg: 149.28 / Max: 317.86 Min: 19.49 / Avg: 154.31 / Max: 321.83 Min: 25.74 / Avg: 158.85 / Max: 377.04 Min: 17.06 / Avg: 161.61 / Max: 378.41 Min: 52.98 / Avg: 188.84 / Max: 288.7 Min: 57.02 / Avg: 215.9 / Max: 374.98 Min: 74.58 / Avg: 223.5 / Max: 376.44 Min: 71.23 / Avg: 226.18 / Max: 377.45 Min: 38.61 / Avg: 248.47 / Max: 653.54 Min: 55.55 / Avg: 279.06 / Max: 655.46 Min: 58.39 / Avg: 299.85 / Max: 772.2 Min: 54.43 / Avg: 323.22 / Max: 766.2 Min: 100.01 / Avg: 325.09 / Max: 579.93 Min: 146.08 / Avg: 400.48 / Max: 754.34 Min: 187.39 / Avg: 424.86 / Max: 754.4 Min: 189.85 / Avg: 429.08 / Max: 754.24
GNU Octave Benchmark This test profile measures how long it takes to complete several reference GNU Octave files via octave-benchmark. GNU Octave is used for numerical computations and is an open-source alternative to MATLAB. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better GNU Octave Benchmark 8.4.0 Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9684X 2P EPYC 9654 Xeon Platinum 8490H Xeon Max 9468 EPYC 9654 2P Xeon Max 9480 Xeon Platinum 8490H 2P EPYC 9754 EPYC 9754 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9480 2P Xeon Max 9468 2P 3 6 9 12 15 SE +/- 0.020, N = 6 SE +/- 0.019, N = 6 SE +/- 0.018, N = 6 SE +/- 0.007, N = 6 SE +/- 0.009, N = 6 SE +/- 0.015, N = 6 SE +/- 0.015, N = 6 SE +/- 0.004, N = 6 SE +/- 0.024, N = 6 SE +/- 0.019, N = 6 SE +/- 0.014, N = 5 SE +/- 0.025, N = 5 SE +/- 0.011, N = 5 SE +/- 0.007, N = 5 SE +/- 0.017, N = 5 SE +/- 0.052, N = 5 SE +/- 0.051, N = 5 SE +/- 0.035, N = 5 SE +/- 0.108, N = 7 SE +/- 0.023, N = 5 SE +/- 0.082, N = 5 SE +/- 0.096, N = 5 6.458 6.856 6.857 6.955 7.528 7.532 7.622 7.645 7.748 8.036 8.801 8.848 8.929 8.948 9.104 9.694 10.243 10.996 11.619 12.245 12.578 12.617 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better GNU Octave Benchmark 8.4.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8592+ Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 Xeon Platinum 8380 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 80 160 240 320 400 Min: 15.49 / Avg: 47.63 / Max: 72.89 Min: 17.25 / Avg: 49.95 / Max: 77.01 Min: 43.6 / Avg: 52.88 / Max: 68.56 Min: 43.45 / Avg: 55 / Max: 74.81 Min: 20.26 / Avg: 67.77 / Max: 105.08 Min: 27.81 / Avg: 84.26 / Max: 135.67 Min: 34.49 / Avg: 95.8 / Max: 154.31 Min: 53.19 / Avg: 97.85 / Max: 151.32 Min: 32.85 / Avg: 108.34 / Max: 202.86 Min: 103.69 / Avg: 114.72 / Max: 127.74 Min: 104.36 / Avg: 116.68 / Max: 139.31 Min: 56.04 / Avg: 133.36 / Max: 217.89 Min: 40.57 / Avg: 161.36 / Max: 222.38 Min: 74.85 / Avg: 164.13 / Max: 237.93 Min: 79.51 / Avg: 166.55 / Max: 232.48 Min: 99.24 / Avg: 180.92 / Max: 246.72 Min: 56.62 / Avg: 189.57 / Max: 270.57 Min: 61.74 / Avg: 205.11 / Max: 301.54 Min: 167.67 / Avg: 296.93 / Max: 375.38 Min: 197.53 / Avg: 371.49 / Max: 456.41 Min: 202.03 / Avg: 373.8 / Max: 446.62
Y-Cruncher Y-Cruncher is a multi-threaded Pi benchmark capable of computing Pi to trillions of digits. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Y-Cruncher 0.8.3 Pi Digits To Calculate: 500M Xeon Platinum 8490H 2P Xeon Platinum 8592+ 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H Xeon Platinum 8592+ Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9480 Xeon Platinum 8380 2P Xeon Max 9468 EPYC 9654 EPYC 9754 EPYC 9754 2P EPYC 9684X EPYC 9654 2P EPYC 9684X 2P Xeon 6780E Xeon Platinum 8380 EPYC 8534P EPYC 8534PN Xeon 6766E 1.1648 2.3296 3.4944 4.6592 5.824 SE +/- 0.002, N = 8 SE +/- 0.004, N = 7 SE +/- 0.074, N = 15 SE +/- 0.064, N = 15 SE +/- 0.003, N = 7 SE +/- 0.008, N = 7 SE +/- 0.004, N = 6 SE +/- 0.004, N = 6 SE +/- 0.023, N = 15 SE +/- 0.006, N = 7 SE +/- 0.010, N = 7 SE +/- 0.003, N = 6 SE +/- 0.005, N = 6 SE +/- 0.015, N = 5 SE +/- 0.003, N = 6 SE +/- 0.003, N = 5 SE +/- 0.007, N = 5 SE +/- 0.007, N = 6 SE +/- 0.003, N = 6 SE +/- 0.006, N = 6 SE +/- 0.008, N = 6 SE +/- 0.005, N = 6 2.364 2.659 2.961 3.110 3.267 3.391 3.411 3.629 3.704 3.724 3.781 3.850 3.968 4.100 4.312 4.329 4.510 4.770 4.874 5.077 5.165 5.177 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Y-Cruncher 0.8.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 16.77 / Avg: 78.49 / Max: 143.84 Min: 16.15 / Avg: 86.13 / Max: 163.32 Min: 43.7 / Avg: 116.04 / Max: 224.06 Min: 20.76 / Avg: 118.84 / Max: 272.84 Min: 43.46 / Avg: 134.29 / Max: 274.09 Min: 27.89 / Avg: 146.25 / Max: 317.02 Min: 31.93 / Avg: 151.06 / Max: 346.25 Min: 52.29 / Avg: 167.23 / Max: 285.41 Min: 32.78 / Avg: 173.58 / Max: 352.72 Min: 57.55 / Avg: 181.21 / Max: 363.23 Min: 70.98 / Avg: 211.16 / Max: 373.06 Min: 74.95 / Avg: 212.09 / Max: 370.6 Min: 104.3 / Avg: 221.86 / Max: 497.5 Min: 106.12 / Avg: 236.59 / Max: 561.92 Min: 37.77 / Avg: 238.41 / Max: 478.35 Min: 81.75 / Avg: 285.35 / Max: 686.94 Min: 53.19 / Avg: 293.51 / Max: 558.3 Min: 97.45 / Avg: 293.71 / Max: 562.12 Min: 66.18 / Avg: 327.67 / Max: 622.14 Min: 164.95 / Avg: 367.73 / Max: 733.59 Min: 189.46 / Avg: 425.51 / Max: 749.34 Min: 197.01 / Avg: 434.63 / Max: 747.05
Embree Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better Embree 4.3 Binary: Pathtracer ISPC - Model: Asian Dragon EPYC 9684X 2P EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9754 EPYC 9684X Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9654 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon 6780E Xeon Platinum 8490H Xeon 6766E Xeon Max 9480 Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 50 100 150 200 250 SE +/- 2.04, N = 8 SE +/- 1.88, N = 7 SE +/- 0.26, N = 8 SE +/- 1.54, N = 8 SE +/- 0.12, N = 7 SE +/- 0.07, N = 7 SE +/- 0.13, N = 7 SE +/- 0.16, N = 7 SE +/- 0.10, N = 6 SE +/- 0.03, N = 6 SE +/- 1.20, N = 15 SE +/- 0.63, N = 6 SE +/- 0.13, N = 6 SE +/- 0.29, N = 6 SE +/- 0.05, N = 5 SE +/- 0.05, N = 5 SE +/- 0.04, N = 5 SE +/- 0.11, N = 5 SE +/- 0.04, N = 5 SE +/- 0.03, N = 5 SE +/- 0.05, N = 5 SE +/- 0.04, N = 4 231.43 208.31 199.61 185.71 156.39 155.12 152.75 151.25 138.85 132.49 129.75 125.75 112.66 99.24 92.88 91.99 84.71 83.49 75.47 75.21 73.91 55.42 Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Embree 4.3 Binary: Pathtracer ISPC - Model: Asian Dragon EPYC 9754 EPYC 9654 EPYC 8534PN EPYC 9684X EPYC 9754 2P EPYC 8534P EPYC 9684X 2P Xeon 6766E Xeon 6780E EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon Max 9480 2P Xeon Max 9468 2P 0.194 0.388 0.582 0.776 0.97 0.862 0.735 0.706 0.662 0.620 0.613 0.564 0.550 0.520 0.518 0.508 0.494 0.457 0.453 0.428 0.323 0.312 0.276 0.258 0.253 0.253 0.246 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Embree 4.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9654 EPYC 9754 Xeon Platinum 8490H Xeon Platinum 8380 Xeon Platinum 8592+ EPYC 9684X Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 140 280 420 560 700 Min: 16.8 / Avg: 104.66 / Max: 145.45 Min: 18.74 / Avg: 122.78 / Max: 171.96 Min: 43.66 / Avg: 153.94 / Max: 217.83 Min: 43.49 / Avg: 178.58 / Max: 261.5 Min: 28.26 / Avg: 180.38 / Max: 335.4 Min: 20.65 / Avg: 181.41 / Max: 329.67 Min: 60.02 / Avg: 215.03 / Max: 377.71 Min: 54.93 / Avg: 219.18 / Max: 290.54 Min: 24.21 / Avg: 228.28 / Max: 376.12 Min: 31.26 / Avg: 234.18 / Max: 410.22 Min: 74.25 / Avg: 267.87 / Max: 377.44 Min: 66.39 / Avg: 273.07 / Max: 377.18 Min: 105.4 / Avg: 303.74 / Max: 488.69 Min: 108.43 / Avg: 333.95 / Max: 567.53 Min: 42.62 / Avg: 335.83 / Max: 642.99 Min: 55.21 / Avg: 358.56 / Max: 646.17 Min: 102.21 / Avg: 383.99 / Max: 581.28 Min: 66.59 / Avg: 393.21 / Max: 754.66 Min: 59.43 / Avg: 410.15 / Max: 817.39 Min: 166.53 / Avg: 472.72 / Max: 754.96 Min: 200.84 / Avg: 511.17 / Max: 752.67 Min: 185.96 / Avg: 511.91 / Max: 753.56
WebP Image Encode This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org MP/s, More Is Better WebP Image Encode 1.2.4 Encode Settings: Quality 100, Highest Compression EPYC 9654 2P EPYC 9654 EPYC 9684X 2P EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8592+ 2P Xeon Max 9480 Xeon Max 9468 Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P EPYC 9754 EPYC 8534PN EPYC 8534P EPYC 9754 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 0.8145 1.629 2.4435 3.258 4.0725 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 6 SE +/- 0.00, N = 5 SE +/- 0.00, N = 5 SE +/- 0.00, N = 5 SE +/- 0.00, N = 5 SE +/- 0.00, N = 5 SE +/- 0.00, N = 5 3.62 3.62 3.62 3.61 3.47 3.46 3.10 3.10 3.09 3.09 3.08 3.07 3.04 3.03 3.03 3.03 2.82 2.81 2.49 2.47 2.23 2.23 1. (CC) gcc options: -fvisibility=hidden -O2 -lm MP/s Per Watt
OpenBenchmarking.org MP/s Per Watt, More Is Better WebP Image Encode 1.2.4 Encode Settings: Quality 100, Highest Compression EPYC 8534PN EPYC 8534P Xeon 6780E Xeon 6766E Xeon Platinum 8592+ EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 Xeon 6780E 2P Xeon 6766E 2P EPYC 9654 2P EPYC 9754 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.0162 0.0324 0.0486 0.0648 0.081 0.072 0.069 0.056 0.051 0.051 0.049 0.048 0.044 0.035 0.033 0.028 0.026 0.023 0.021 0.021 0.020 0.019 0.019 0.017 0.014 0.011 0.011 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better WebP Image Encode 1.2.4 CPU Power Consumption Monitor EPYC 8534PN Xeon 6766E Xeon 6780E EPYC 8534P EPYC 9754 Xeon Platinum 8592+ EPYC 9654 EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8490H Xeon 6766E 2P Xeon 6780E 2P Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P Xeon Platinum 8380 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 70 140 210 280 350 Min: 16.53 / Avg: 42.26 / Max: 55.05 Min: 43.59 / Avg: 43.93 / Max: 46.02 Min: 43.35 / Avg: 44.16 / Max: 45.33 Min: 16.45 / Avg: 44.16 / Max: 57.18 Min: 20.53 / Avg: 61.84 / Max: 81.06 Min: 34.25 / Avg: 68.3 / Max: 101.37 Min: 28.98 / Avg: 75.95 / Max: 101.21 Min: 32.65 / Avg: 82.63 / Max: 115.67 Min: 50.09 / Avg: 86.28 / Max: 102.41 Min: 59.48 / Avg: 88.57 / Max: 151.52 Min: 103.85 / Avg: 107.03 / Max: 109.06 Min: 105.24 / Avg: 107.9 / Max: 111.31 Min: 70.24 / Avg: 111.05 / Max: 144.04 Min: 81.36 / Avg: 119.53 / Max: 163.62 Min: 39.42 / Avg: 149.38 / Max: 197.42 Min: 100.4 / Avg: 169.36 / Max: 210.51 Min: 56.66 / Avg: 176 / Max: 238.56 Min: 74.68 / Avg: 186.13 / Max: 288.17 Min: 63.07 / Avg: 192.89 / Max: 262.1 Min: 148.59 / Avg: 222.03 / Max: 316.69 Min: 200.76 / Avg: 274.15 / Max: 348.74 Min: 205.94 / Avg: 277 / Max: 371.07
Embree Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better Embree 4.3 Binary: Pathtracer ISPC - Model: Crown EPYC 9684X 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon 6780E 2P EPYC 9684X EPYC 9754 Xeon Platinum 8490H 2P Xeon 6766E 2P Xeon Max 9480 2P EPYC 9654 Xeon Max 9468 2P Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon 6780E Xeon 6766E Xeon Platinum 8490H Xeon Max 9480 EPYC 8534P Xeon Max 9468 EPYC 8534PN Xeon Platinum 8380 50 100 150 200 250 SE +/- 1.23, N = 8 SE +/- 1.04, N = 7 SE +/- 0.43, N = 7 SE +/- 0.07, N = 7 SE +/- 0.14, N = 6 SE +/- 0.20, N = 7 SE +/- 0.64, N = 6 SE +/- 0.12, N = 6 SE +/- 0.11, N = 6 SE +/- 0.48, N = 6 SE +/- 0.08, N = 6 SE +/- 0.14, N = 6 SE +/- 0.05, N = 5 SE +/- 0.08, N = 5 SE +/- 0.02, N = 5 SE +/- 0.05, N = 5 SE +/- 0.07, N = 5 SE +/- 0.02, N = 4 SE +/- 0.04, N = 4 SE +/- 0.05, N = 4 SE +/- 0.07, N = 4 SE +/- 0.07, N = 3 214.02 190.08 176.21 153.16 133.42 130.79 126.15 124.71 121.43 116.61 110.53 107.63 82.97 82.92 81.83 73.96 70.05 65.33 62.19 58.42 58.28 43.67 Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better Embree 4.3 Binary: Pathtracer ISPC - Model: Crown EPYC 9754 EPYC 9754 2P EPYC 9684X EPYC 8534PN EPYC 9654 EPYC 9684X 2P EPYC 8534P EPYC 9654 2P Xeon 6766E Xeon 6780E Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Platinum 8380 2P Xeon Max 9468 2P Xeon Platinum 8380 0.1415 0.283 0.4245 0.566 0.7075 0.629 0.532 0.522 0.517 0.499 0.499 0.482 0.468 0.422 0.381 0.377 0.373 0.363 0.331 0.258 0.252 0.232 0.226 0.206 0.205 0.205 0.192 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Embree 4.3 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E EPYC 9754 Xeon 6780E EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon 6766E 2P EPYC 9754 2P Xeon 6780E 2P EPYC 9654 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 16.58 / Avg: 112.82 / Max: 146.36 Min: 17.65 / Avg: 128.92 / Max: 171.09 Min: 43.6 / Avg: 175.27 / Max: 236.71 Min: 20.38 / Avg: 200.39 / Max: 335.38 Min: 43.42 / Avg: 214.8 / Max: 302.74 Min: 27.85 / Avg: 221.71 / Max: 341.98 Min: 53.87 / Avg: 227.2 / Max: 288.12 Min: 32.54 / Avg: 250.47 / Max: 410.99 Min: 26.17 / Avg: 250.53 / Max: 384.62 Min: 60.79 / Avg: 271.03 / Max: 377.88 Min: 75.4 / Avg: 281.67 / Max: 377.64 Min: 71.98 / Avg: 284.3 / Max: 376.72 Min: 104.47 / Avg: 322.07 / Max: 484.23 Min: 39.31 / Avg: 357.16 / Max: 678.39 Min: 105.86 / Avg: 357.3 / Max: 564.3 Min: 58.37 / Avg: 376.52 / Max: 669.54 Min: 102.04 / Avg: 404.97 / Max: 574.07 Min: 65.93 / Avg: 422.5 / Max: 748.67 Min: 60.99 / Avg: 428.94 / Max: 819.87 Min: 171.82 / Avg: 493.94 / Max: 734.13 Min: 178.12 / Avg: 516.97 / Max: 755.2 Min: 192.27 / Avg: 525.15 / Max: 751.36
ASTC Encoder ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 4.7 Preset: Medium Xeon Platinum 8592+ 2P EPYC 9754 EPYC 9684X 2P EPYC 9684X EPYC 9654 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P EPYC 9754 2P EPYC 9654 Xeon Max 9468 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8490H Xeon Platinum 8592+ Xeon 6780E EPYC 8534P Xeon 6766E Xeon Max 9480 EPYC 8534PN Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8380 120 240 360 480 600 SE +/- 1.56, N = 7 SE +/- 0.39, N = 7 SE +/- 3.11, N = 7 SE +/- 0.63, N = 7 SE +/- 3.54, N = 6 SE +/- 2.79, N = 6 SE +/- 0.77, N = 6 SE +/- 0.52, N = 6 SE +/- 0.20, N = 7 SE +/- 0.52, N = 6 SE +/- 2.01, N = 5 SE +/- 1.77, N = 5 SE +/- 0.16, N = 6 SE +/- 1.17, N = 6 SE +/- 0.13, N = 6 SE +/- 0.07, N = 6 SE +/- 0.18, N = 5 SE +/- 0.08, N = 6 SE +/- 0.07, N = 6 SE +/- 0.03, N = 6 SE +/- 2.26, N = 5 SE +/- 0.32, N = 4 550.02 527.80 508.09 500.77 491.53 484.55 483.10 462.24 449.11 441.65 363.38 338.15 333.99 327.78 314.62 294.62 284.46 284.33 273.67 247.30 216.97 126.44 1. (CXX) g++ options: -O3 -flto -pthread MT/s Per Watt
OpenBenchmarking.org MT/s Per Watt, More Is Better ASTC Encoder 4.7 Preset: Medium EPYC 9754 EPYC 8534PN EPYC 8534P EPYC 9654 EPYC 9684X Xeon 6766E Xeon 6780E EPYC 9754 2P EPYC 9654 2P Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9480 2P Xeon Max 9468 Xeon Max 9468 2P Xeon Platinum 8380 2P Xeon Platinum 8380 1.1363 2.2726 3.4089 4.5452 5.6815 5.050 3.989 3.926 3.553 3.488 3.191 3.076 2.377 2.099 2.084 2.048 2.012 1.983 1.966 1.923 1.414 1.373 1.215 1.167 1.094 0.741 0.720 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ASTC Encoder 4.7 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 Xeon 6780E 2P EPYC 9754 2P Xeon Max 9480 Xeon Max 9468 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 130 260 390 520 650 Min: 15.15 / Avg: 68.61 / Max: 143.04 Min: 16.05 / Avg: 75.04 / Max: 166.06 Min: 43.61 / Avg: 89.14 / Max: 212.79 Min: 43.4 / Avg: 102.28 / Max: 269.1 Min: 20.85 / Avg: 104.51 / Max: 314.72 Min: 27.98 / Avg: 126.4 / Max: 324.02 Min: 32.4 / Avg: 143.57 / Max: 393.87 Min: 103.94 / Avg: 162.27 / Max: 347.6 Min: 26.56 / Avg: 162.88 / Max: 364.95 Min: 60.47 / Avg: 173.64 / Max: 368.66 Min: 49.69 / Avg: 175.72 / Max: 288.95 Min: 105.97 / Avg: 177.44 / Max: 411 Min: 38.51 / Avg: 194.48 / Max: 419.92 Min: 46.45 / Avg: 207.12 / Max: 368.04 Min: 73.51 / Avg: 211.87 / Max: 372.54 Min: 57.21 / Avg: 234.22 / Max: 515.92 Min: 60.92 / Avg: 258.49 / Max: 596.25 Min: 75.77 / Avg: 277.37 / Max: 691.02 Min: 94.9 / Avg: 291.19 / Max: 532.11 Min: 149.65 / Avg: 342.67 / Max: 671.28 Min: 188.21 / Avg: 397.48 / Max: 724.85 Min: 202.14 / Avg: 403.63 / Max: 727.09
SVT-AV1 This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.0 Encoder Mode: Preset 12 - Input: Bosphorus 4K EPYC 9654 Xeon Platinum 8592+ EPYC 9654 2P EPYC 9754 EPYC 9754 2P EPYC 9684X Xeon Platinum 8592+ 2P EPYC 8534P Xeon Platinum 8490H EPYC 9684X 2P EPYC 8534PN Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon Max 9480 2P Xeon 6780E Xeon Max 9468 2P Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 40 80 120 160 200 SE +/- 0.80, N = 7 SE +/- 0.20, N = 7 SE +/- 1.88, N = 15 SE +/- 1.32, N = 6 SE +/- 1.41, N = 6 SE +/- 0.43, N = 7 SE +/- 0.42, N = 7 SE +/- 0.71, N = 6 SE +/- 0.41, N = 7 SE +/- 0.88, N = 6 SE +/- 0.78, N = 6 SE +/- 0.25, N = 7 SE +/- 0.40, N = 7 SE +/- 0.61, N = 7 SE +/- 1.28, N = 6 SE +/- 0.48, N = 6 SE +/- 0.97, N = 6 SE +/- 0.15, N = 6 SE +/- 0.83, N = 5 SE +/- 1.07, N = 6 SE +/- 0.08, N = 6 SE +/- 0.88, N = 5 178.52 177.86 171.22 169.23 168.78 165.75 165.50 165.09 164.25 163.80 157.36 156.52 151.72 146.10 140.76 134.07 127.55 126.59 122.09 117.68 115.35 108.50 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq Frames Per Second Per Watt
OpenBenchmarking.org Frames Per Second Per Watt, More Is Better SVT-AV1 2.0 Encoder Mode: Preset 12 - Input: Bosphorus 4K EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6780E EPYC 9654 Xeon 6766E EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Platinum 8380 EPYC 9754 2P Xeon 6780E 2P Xeon 6766E 2P EPYC 9654 2P Xeon Max 9480 Xeon Max 9468 EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 0.6399 1.2798 1.9197 2.5596 3.1995 2.844 2.784 2.019 1.741 1.714 1.703 1.399 1.363 1.097 1.054 1.001 0.843 0.833 0.829 0.812 0.784 0.709 0.675 0.638 0.500 0.318 0.306 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better SVT-AV1 2.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8380 Xeon 6766E 2P Xeon Platinum 8592+ Xeon 6780E 2P Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 100 200 300 400 500 Min: 14.97 / Avg: 55.32 / Max: 108.24 Min: 16.3 / Avg: 59.31 / Max: 124.91 Min: 43.6 / Avg: 67.73 / Max: 107.71 Min: 43.49 / Avg: 72.69 / Max: 125.26 Min: 21.46 / Avg: 83.8 / Max: 173.31 Min: 25.96 / Avg: 104.16 / Max: 214.25 Min: 28.14 / Avg: 118.49 / Max: 228.43 Min: 49.17 / Avg: 127.22 / Max: 220.81 Min: 104.27 / Avg: 130.28 / Max: 174.4 Min: 27.81 / Avg: 130.54 / Max: 271.31 Min: 105.73 / Avg: 139.65 / Max: 205.54 Min: 58.81 / Avg: 149.76 / Max: 276.69 Min: 39.75 / Avg: 168.64 / Max: 285.91 Min: 78.53 / Avg: 186.44 / Max: 299.02 Min: 70.86 / Avg: 186.8 / Max: 304.11 Min: 57.32 / Avg: 206.47 / Max: 350.04 Min: 98.84 / Avg: 220.59 / Max: 378.11 Min: 64.22 / Avg: 231.05 / Max: 380.24 Min: 72.44 / Avg: 245.05 / Max: 463.8 Min: 162.99 / Avg: 313.03 / Max: 463.03 Min: 206.55 / Avg: 399.09 / Max: 553.52 Min: 192.02 / Avg: 400.5 / Max: 567.38
libavif avifenc This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 1.0 Encoder Speed: 6, Lossless Xeon Platinum 8592+ 2P EPYC 9684X Xeon Platinum 8592+ EPYC 9654 EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 EPYC 9754 EPYC 9754 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 8534P EPYC 8534PN Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 3 6 9 12 15 SE +/- 0.031, N = 7 SE +/- 0.014, N = 7 SE +/- 0.025, N = 7 SE +/- 0.010, N = 7 SE +/- 0.020, N = 7 SE +/- 0.015, N = 7 SE +/- 0.022, N = 7 SE +/- 0.013, N = 7 SE +/- 0.033, N = 6 SE +/- 0.045, N = 6 SE +/- 0.007, N = 6 SE +/- 0.025, N = 6 SE +/- 0.035, N = 6 SE +/- 0.027, N = 6 SE +/- 0.009, N = 6 SE +/- 0.011, N = 6 SE +/- 0.027, N = 6 SE +/- 0.016, N = 6 SE +/- 0.010, N = 6 SE +/- 0.012, N = 5 SE +/- 0.038, N = 5 SE +/- 0.054, N = 5 5.613 5.757 5.784 5.799 5.846 5.950 6.172 6.278 6.605 6.836 6.853 6.986 7.038 7.121 7.182 7.301 7.388 8.201 8.359 9.188 9.351 9.960 1. (CXX) g++ options: -O3 -fPIC -lm CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better libavif avifenc 1.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon 6766E 2P Xeon Platinum 8592+ Xeon 6780E 2P Xeon Platinum 8380 Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9480 Xeon Max 9468 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 120 240 360 480 600 Min: 15.05 / Avg: 53.07 / Max: 124.5 Min: 15.99 / Avg: 55.18 / Max: 149.59 Min: 43.56 / Avg: 59.76 / Max: 149.48 Min: 43.39 / Avg: 64.09 / Max: 183.29 Min: 20.4 / Avg: 80.51 / Max: 234.12 Min: 27.48 / Avg: 100.34 / Max: 306.84 Min: 29.45 / Avg: 115.08 / Max: 342.75 Min: 53.34 / Avg: 123.17 / Max: 220.64 Min: 26.64 / Avg: 128.43 / Max: 374.81 Min: 106.3 / Avg: 128.53 / Max: 249.32 Min: 48.9 / Avg: 132.08 / Max: 290.31 Min: 61.6 / Avg: 158.79 / Max: 375.47 Min: 39.73 / Avg: 165.83 / Max: 352.71 Min: 69.76 / Avg: 191.83 / Max: 376.91 Min: 74.59 / Avg: 193.8 / Max: 362.37 Min: 51.5 / Avg: 198.44 / Max: 459.76 Min: 95.62 / Avg: 213.07 / Max: 530.05 Min: 61.29 / Avg: 220.4 / Max: 507.64 Min: 74.34 / Avg: 248.3 / Max: 649.52 Min: 143.68 / Avg: 313.6 / Max: 692.26 Min: 196.47 / Avg: 378.09 / Max: 677.18 Min: 197.54 / Avg: 390.59 / Max: 684.36
ACES DGEMM This is a multi-threaded DGEMM benchmark. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org GFLOP/s, More Is Better ACES DGEMM 1.0 Sustained Floating-Point Rate EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9684X EPYC 9754 Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9654 Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Platinum 8490H Xeon Max 9480 EPYC 8534P EPYC 8534PN Xeon Max 9468 Xeon 6780E Xeon 6766E Xeon Platinum 8380 20 40 60 80 100 SE +/- 0.19, N = 10 SE +/- 0.39, N = 9 SE +/- 0.10, N = 9 SE +/- 0.11, N = 8 SE +/- 0.28, N = 7 SE +/- 0.40, N = 7 SE +/- 0.17, N = 7 SE +/- 0.04, N = 7 SE +/- 0.06, N = 7 SE +/- 0.22, N = 7 SE +/- 0.04, N = 6 SE +/- 0.03, N = 6 SE +/- 0.08, N = 6 SE +/- 0.09, N = 5 SE +/- 0.09, N = 5 SE +/- 0.09, N = 5 SE +/- 0.16, N = 5 SE +/- 0.10, N = 5 SE +/- 0.04, N = 4 SE +/- 0.01, N = 4 SE +/- 0.01, N = 4 SE +/- 0.01, N = 3 84.69 78.50 68.87 58.25 45.97 45.25 44.97 44.67 39.35 39.09 35.20 31.48 29.14 24.97 23.54 22.54 21.82 21.00 19.77 17.85 15.94 13.49 1. (CC) gcc options: -O3 -march=native -fopenmp GFLOP/s Per Watt
OpenBenchmarking.org GFLOP/s Per Watt, More Is Better ACES DGEMM 1.0 Sustained Floating-Point Rate EPYC 9754 2P EPYC 9754 EPYC 9654 2P EPYC 8534PN EPYC 9684X 2P EPYC 9654 EPYC 8534P EPYC 9684X Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon 6766E 2P Xeon 6780E 2P Xeon 6766E Xeon 6780E Xeon Platinum 8490H 2P Xeon Platinum 8490H Xeon Max 9480 2P Xeon Max 9480 Xeon Max 9468 2P Xeon Max 9468 Xeon Platinum 8380 2P Xeon Platinum 8380 0.0657 0.1314 0.1971 0.2628 0.3285 0.292 0.252 0.228 0.220 0.215 0.214 0.198 0.196 0.149 0.121 0.114 0.112 0.107 0.105 0.101 0.093 0.090 0.081 0.078 0.071 0.063 0.059 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better ACES DGEMM 1.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 Xeon 6766E 2P EPYC 9754 2P EPYC 9654 2P Xeon 6780E 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 16.53 / Avg: 95.47 / Max: 135.35 Min: 17.88 / Avg: 109.94 / Max: 158.28 Min: 43.57 / Avg: 148.35 / Max: 183.52 Min: 43.36 / Avg: 170.6 / Max: 222.09 Min: 22.22 / Avg: 178.2 / Max: 342.37 Min: 28.2 / Avg: 182.5 / Max: 306.04 Min: 52.19 / Avg: 226.98 / Max: 281.28 Min: 31.49 / Avg: 230.52 / Max: 409.55 Min: 23.09 / Avg: 239.87 / Max: 363.28 Min: 53.98 / Avg: 252.19 / Max: 377.52 Min: 75.8 / Avg: 276.9 / Max: 361.7 Min: 73.75 / Avg: 277.15 / Max: 377.23 Min: 104.19 / Avg: 277.33 / Max: 390.75 Min: 40.76 / Avg: 290.43 / Max: 650.39 Min: 56.61 / Avg: 302.67 / Max: 601.18 Min: 106.31 / Avg: 312.9 / Max: 465.14 Min: 63.45 / Avg: 364.55 / Max: 774.44 Min: 68.08 / Avg: 390.95 / Max: 737.84 Min: 99.45 / Avg: 398.61 / Max: 554.79 Min: 164.24 / Avg: 453.94 / Max: 741.45 Min: 194.98 / Avg: 496.59 / Max: 754.16 Min: 193.63 / Avg: 507 / Max: 747.89
libavif avifenc This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 1.0 Encoder Speed: 10, Lossless Xeon Platinum 8592+ Xeon Platinum 8592+ 2P EPYC 9684X EPYC 9654 EPYC 9654 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9754 EPYC 8534PN EPYC 8534P EPYC 9754 2P Xeon Platinum 8380 Xeon Platinum 8380 2P Xeon 6780E Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 2 4 6 8 10 SE +/- 0.008, N = 8 SE +/- 0.015, N = 8 SE +/- 0.002, N = 8 SE +/- 0.006, N = 8 SE +/- 0.006, N = 8 SE +/- 0.015, N = 7 SE +/- 0.005, N = 7 SE +/- 0.006, N = 7 SE +/- 0.003, N = 7 SE +/- 0.014, N = 7 SE +/- 0.028, N = 7 SE +/- 0.019, N = 7 SE +/- 0.004, N = 7 SE +/- 0.010, N = 7 SE +/- 0.006, N = 7 SE +/- 0.006, N = 7 SE +/- 0.008, N = 7 SE +/- 0.011, N = 7 SE +/- 0.003, N = 6 SE +/- 0.009, N = 6 SE +/- 0.016, N = 6 SE +/- 0.021, N = 6 4.456 4.579 4.770 4.792 4.829 4.867 4.916 4.923 4.943 5.074 5.311 5.317 5.662 5.734 5.735 5.740 5.906 5.919 6.849 7.507 7.603 8.094 1. (CXX) g++ options: -O3 -fPIC -lm CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better libavif avifenc 1.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8380 Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9480 Xeon Max 9468 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 100 200 300 400 500 Min: 15.19 / Avg: 43.89 / Max: 93.61 Min: 15.89 / Avg: 45.79 / Max: 103.55 Min: 43.59 / Avg: 50.67 / Max: 89.41 Min: 43.45 / Avg: 52.35 / Max: 100.96 Min: 20.59 / Avg: 64.93 / Max: 140.09 Min: 27.35 / Avg: 79.69 / Max: 177.97 Min: 31.1 / Avg: 89.96 / Max: 203.77 Min: 26.86 / Avg: 95.6 / Max: 261.18 Min: 49.47 / Avg: 100.66 / Max: 234.57 Min: 103.88 / Avg: 113.94 / Max: 151.55 Min: 106.05 / Avg: 116.77 / Max: 170.77 Min: 58.72 / Avg: 130.94 / Max: 271.24 Min: 39.49 / Avg: 147.26 / Max: 254.64 Min: 72.94 / Avg: 159.29 / Max: 307.76 Min: 73.28 / Avg: 159.39 / Max: 303.29 Min: 54.59 / Avg: 171.46 / Max: 320.72 Min: 97.81 / Avg: 180.38 / Max: 349.14 Min: 59.66 / Avg: 192.17 / Max: 352.61 Min: 76.28 / Avg: 204.47 / Max: 438.17 Min: 134.59 / Avg: 276.23 / Max: 451.3 Min: 197.84 / Avg: 346.41 / Max: 545.23 Min: 198.99 / Avg: 348.01 / Max: 540.22
NAS Parallel Benchmarks NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: MG.C EPYC 9754 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8490H 2P EPYC 9684X EPYC 9754 EPYC 9654 Xeon Platinum 8380 2P Xeon Platinum 8592+ Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8490H EPYC 8534P EPYC 8534PN Xeon Platinum 8380 50K 100K 150K 200K 250K SE +/- 1776.37, N = 11 SE +/- 7207.31, N = 15 SE +/- 1422.07, N = 7 SE +/- 2707.72, N = 14 SE +/- 1415.96, N = 9 SE +/- 1298.67, N = 9 SE +/- 1262.04, N = 9 SE +/- 1372.89, N = 15 SE +/- 884.47, N = 9 SE +/- 1153.71, N = 15 SE +/- 233.83, N = 9 SE +/- 406.29, N = 9 SE +/- 382.59, N = 10 SE +/- 436.97, N = 10 SE +/- 152.92, N = 9 SE +/- 167.65, N = 8 SE +/- 301.54, N = 8 SE +/- 53.56, N = 9 245447.72 244738.81 220893.78 194942.32 172614.10 171436.47 163333.21 134994.14 125567.07 116725.04 110742.15 107321.89 86407.46 85910.95 83448.70 63930.64 62449.55 57317.66 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6 Total Mop/s Per Watt
OpenBenchmarking.org Total Mop/s Per Watt, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: MG.C EPYC 9754 EPYC 9684X EPYC 9654 EPYC 9754 2P EPYC 8534PN EPYC 9684X 2P EPYC 8534P Xeon Platinum 8592+ EPYC 9654 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H Xeon Max 9468 Xeon Platinum 8490H 2P Xeon Max 9480 Xeon Max 9468 2P Xeon Max 9480 2P Xeon Platinum 8380 2P Xeon Platinum 8380 300 600 900 1200 1500 1265.55 1119.20 1070.68 988.65 953.72 928.22 906.46 858.24 823.46 757.48 517.49 499.08 496.64 480.22 476.05 454.47 446.48 378.76 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better NAS Parallel Benchmarks 3.4 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 16.49 / Avg: 65.48 / Max: 127.89 Min: 17.11 / Avg: 70.53 / Max: 141.22 Min: 21.58 / Avg: 99.22 / Max: 284.09 Min: 27.57 / Avg: 109.02 / Max: 286.49 Min: 29.67 / Avg: 120.62 / Max: 340.18 Min: 25.53 / Avg: 125.05 / Max: 346.36 Min: 52.21 / Avg: 151.33 / Max: 280.05 Min: 59.14 / Avg: 161.26 / Max: 333.63 Min: 72.77 / Avg: 172.14 / Max: 347.7 Min: 73.21 / Avg: 179.93 / Max: 359.94 Min: 55.9 / Avg: 236.74 / Max: 614.68 Min: 99.93 / Avg: 248.03 / Max: 546.61 Min: 40.32 / Avg: 248.27 / Max: 531.63 Min: 62.46 / Avg: 263.66 / Max: 710.66 Min: 70.18 / Avg: 291.62 / Max: 590.65 Min: 142.29 / Avg: 328.87 / Max: 698.64 Min: 183.3 / Avg: 362.6 / Max: 714.29 Min: 193.92 / Avg: 377.22 / Max: 729.73
srsRAN Project srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Mbps, More Is Better srsRAN Project 23.10.1-20240325 Test: PDSCH Processor Benchmark, Throughput Thread Xeon Platinum 8592+ 2P Xeon Platinum 8592+ Xeon Max 9480 Xeon Max 9468 EPYC 9654 EPYC 9684X Xeon Platinum 8490H Xeon Platinum 8490H 2P EPYC 9654 2P EPYC 9684X 2P Xeon Max 9480 2P Xeon Max 9468 2P EPYC 9754 EPYC 9754 2P EPYC 8534PN EPYC 8534P Xeon Platinum 8380 2P Xeon Platinum 8380 Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 160 320 480 640 800 SE +/- 1.89, N = 8 SE +/- 1.34, N = 8 SE +/- 1.27, N = 8 SE +/- 1.64, N = 8 SE +/- 1.71, N = 8 SE +/- 3.13, N = 8 SE +/- 1.17, N = 8 SE +/- 2.57, N = 8 SE +/- 3.72, N = 8 SE +/- 5.52, N = 8 SE +/- 5.02, N = 8 SE +/- 3.86, N = 8 SE +/- 0.80, N = 7 SE +/- 4.48, N = 7 SE +/- 1.24, N = 7 SE +/- 3.14, N = 7 SE +/- 2.08, N = 7 SE +/- 3.57, N = 7 SE +/- 0.24, N = 5 SE +/- 0.66, N = 5 SE +/- 0.32, N = 5 SE +/- 0.32, N = 5 755.5 742.9 705.5 702.6 684.7 680.7 680.2 675.5 673.4 664.7 663.8 651.9 596.2 576.8 563.6 557.9 490.3 484.8 333.2 323.7 299.4 293.1 1. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -fno-trapping-math -fno-math-errno -ldl Mbps Per Watt
OpenBenchmarking.org Mbps Per Watt, More Is Better srsRAN Project 23.10.1-20240325 Test: PDSCH Processor Benchmark, Throughput Thread EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8592+ EPYC 9684X Xeon 6780E Xeon 6766E Xeon Platinum 8490H Xeon Platinum 8380 Xeon Max 9468 Xeon Max 9480 EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 4 8 12 16 20 14.524 13.548 10.675 9.795 8.929 8.628 7.249 6.807 6.116 5.866 5.085 4.788 4.200 4.162 4.074 3.639 2.998 2.979 2.726 2.598 2.065 2.035 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better srsRAN Project 23.10.1-20240325 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8592+ Xeon 6766E 2P Xeon 6780E 2P Xeon Platinum 8490H EPYC 9754 2P Xeon Max 9468 Xeon Max 9480 EPYC 9654 2P Xeon Platinum 8380 2P EPYC 9684X 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 70 140 210 280 350 Min: 13.69 / Avg: 38.81 / Max: 55.52 Min: 18.39 / Avg: 41.18 / Max: 58.06 Min: 43.56 / Avg: 43.98 / Max: 45.49 Min: 43.52 / Avg: 45.97 / Max: 48.12 Min: 20.14 / Avg: 55.85 / Max: 81.57 Min: 28.33 / Avg: 69.9 / Max: 103.51 Min: 30.33 / Avg: 78.89 / Max: 117.61 Min: 51.21 / Avg: 82.65 / Max: 104.34 Min: 24.93 / Avg: 83.2 / Max: 132.66 Min: 103.98 / Avg: 107.54 / Max: 109.81 Min: 104.77 / Avg: 108.64 / Max: 112.06 Min: 62.63 / Avg: 111.22 / Max: 152.68 Min: 38.06 / Avg: 137.34 / Max: 200.95 Min: 73.91 / Avg: 138.16 / Max: 195.36 Min: 85.93 / Avg: 147.34 / Max: 200.38 Min: 56.59 / Avg: 161.78 / Max: 243.41 Min: 98.94 / Avg: 163.54 / Max: 216.84 Min: 63.99 / Avg: 182.65 / Max: 268.9 Min: 80.15 / Avg: 185.44 / Max: 287.05 Min: 161.37 / Avg: 259.97 / Max: 337.8 Min: 204.49 / Avg: 320.28 / Max: 415.75 Min: 199.86 / Avg: 321.38 / Max: 411.1
LAMMPS Molecular Dynamics Simulator LAMMPS is a classical molecular dynamics code, and an acronym for Large-scale Atomic/Molecular Massively Parallel Simulator. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ns/day, More Is Better LAMMPS Molecular Dynamics Simulator 23Jun2022 Model: Rhodopsin Protein Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Max 9480 2P EPYC 9754 2P EPYC 9654 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P EPYC 9754 EPYC 9684X EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8380 2P Xeon Max 9480 Xeon Platinum 8490H Xeon Max 9468 EPYC 8534P EPYC 8534PN Xeon Platinum 8380 12 24 36 48 60 SE +/- 0.97, N = 15 SE +/- 0.43, N = 7 SE +/- 0.09, N = 8 SE +/- 0.24, N = 5 SE +/- 0.39, N = 7 SE +/- 0.11, N = 9 SE +/- 0.33, N = 8 SE +/- 0.29, N = 10 SE +/- 0.35, N = 15 SE +/- 0.14, N = 11 SE +/- 0.29, N = 15 SE +/- 0.53, N = 15 SE +/- 0.04, N = 11 SE +/- 0.36, N = 15 SE +/- 0.05, N = 11 SE +/- 0.29, N = 15 SE +/- 0.30, N = 15 SE +/- 0.28, N = 15 55.50 53.77 52.24 51.63 50.12 47.51 42.28 40.39 40.00 36.15 34.30 33.54 30.84 28.72 27.63 24.25 21.94 21.16 1. (CXX) g++ options: -O3 -lm -ldl ns/day Per Watt
OpenBenchmarking.org ns/day Per Watt, More Is Better LAMMPS Molecular Dynamics Simulator 23Jun2022 Model: Rhodopsin Protein EPYC 8534P EPYC 8534PN EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon Max 9480 Xeon Platinum 8490H EPYC 9754 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8380 Xeon Max 9468 EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8490H 2P 0.104 0.208 0.312 0.416 0.52 0.462 0.451 0.451 0.372 0.358 0.341 0.209 0.209 0.204 0.196 0.189 0.186 0.184 0.179 0.148 0.136 0.129 0.119 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better LAMMPS Molecular Dynamics Simulator 23Jun2022 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 EPYC 9654 Xeon Platinum 8592+ EPYC 9684X Xeon Platinum 8380 Xeon Platinum 8490H Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2P EPYC 9754 2P EPYC 9654 2P Xeon Platinum 8592+ 2P EPYC 9684X 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 120 240 360 480 600 Min: 15.52 / Avg: 48.67 / Max: 132.86 Min: 16.36 / Avg: 52.44 / Max: 142.05 Min: 20.56 / Avg: 89.57 / Max: 288.34 Min: 26.87 / Avg: 97.2 / Max: 234.7 Min: 22.32 / Avg: 100.63 / Max: 327.13 Min: 27.48 / Avg: 111.71 / Max: 349.02 Min: 51.92 / Avg: 113.55 / Max: 290.07 Min: 59.51 / Avg: 137.71 / Max: 356.46 Min: 73.75 / Avg: 147.45 / Max: 350.72 Min: 74.62 / Avg: 150.09 / Max: 352.27 Min: 99.81 / Avg: 226.26 / Max: 547.72 Min: 41.5 / Avg: 252.99 / Max: 566.7 Min: 56.67 / Avg: 265.38 / Max: 605.84 Min: 63.52 / Avg: 283.35 / Max: 590.85 Min: 60.69 / Avg: 300.13 / Max: 690.91 Min: 158.94 / Avg: 354.59 / Max: 652.14 Min: 188.46 / Avg: 368.48 / Max: 640.19 Min: 185.06 / Avg: 384.54 / Max: 575.71
libavif avifenc This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 1.0 Encoder Speed: 6 EPYC 9684X EPYC 9654 Xeon Platinum 8592+ 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8592+ Xeon Platinum 8490H 2P Xeon Platinum 8490H EPYC 9754 EPYC 9754 2P EPYC 8534P Xeon Platinum 8380 2P Xeon Max 9480 EPYC 8534PN Xeon Max 9468 Xeon Max 9480 2P Xeon Max 9468 2P Xeon 6780E Xeon Platinum 8380 Xeon 6766E Xeon 6780E 2P Xeon 6766E 2P 1.0865 2.173 3.2595 4.346 5.4325 SE +/- 0.005, N = 10 SE +/- 0.011, N = 10 SE +/- 0.009, N = 10 SE +/- 0.008, N = 10 SE +/- 0.005, N = 9 SE +/- 0.016, N = 9 SE +/- 0.008, N = 9 SE +/- 0.022, N = 9 SE +/- 0.002, N = 9 SE +/- 0.026, N = 9 SE +/- 0.016, N = 9 SE +/- 0.023, N = 15 SE +/- 0.011, N = 9 SE +/- 0.011, N = 9 SE +/- 0.017, N = 9 SE +/- 0.022, N = 9 SE +/- 0.029, N = 8 SE +/- 0.004, N = 8 SE +/- 0.010, N = 8 SE +/- 0.003, N = 8 SE +/- 0.034, N = 8 SE +/- 0.023, N = 8 2.723 2.756 2.805 2.810 2.890 2.971 3.044 3.169 3.251 3.341 3.484 3.528 3.550 3.623 3.730 3.738 3.776 3.872 4.093 4.243 4.486 4.829 1. (CXX) g++ options: -O3 -fPIC -lm CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better libavif avifenc 1.0 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8592+ Xeon 6766E 2P Xeon Platinum 8380 Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8490H EPYC 9654 2P Xeon Max 9468 Xeon Max 9480 EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 120 240 360 480 600 Min: 15.57 / Avg: 49.3 / Max: 126.89 Min: 15.86 / Avg: 54.43 / Max: 150.92 Min: 43.55 / Avg: 64.88 / Max: 147.13 Min: 43.39 / Avg: 69.56 / Max: 179.96 Min: 18.49 / Avg: 76.34 / Max: 238.8 Min: 26.55 / Avg: 92.59 / Max: 299.06 Min: 29.87 / Avg: 106.31 / Max: 349.7 Min: 27.77 / Avg: 125.51 / Max: 374.47 Min: 103.77 / Avg: 129.25 / Max: 219.98 Min: 49.52 / Avg: 131.52 / Max: 290.38 Min: 106.17 / Avg: 135.03 / Max: 247.79 Min: 35.83 / Avg: 144.55 / Max: 349.42 Min: 59.37 / Avg: 149.47 / Max: 376.13 Min: 54.93 / Avg: 178.85 / Max: 447.63 Min: 73.48 / Avg: 180.42 / Max: 357.76 Min: 71.59 / Avg: 183.17 / Max: 376.96 Min: 59.74 / Avg: 193.24 / Max: 500.52 Min: 96.12 / Avg: 215.89 / Max: 534.91 Min: 66.73 / Avg: 224.06 / Max: 669.89 Min: 141.74 / Avg: 294.92 / Max: 681.8 Min: 196.21 / Avg: 370.24 / Max: 687.54 Min: 194.63 / Avg: 375.42 / Max: 682.97
Primesieve Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Primesieve 12.1 Length: 1e12 EPYC 9684X 2P EPYC 9654 2P EPYC 9754 2P EPYC 9754 EPYC 9684X EPYC 9654 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon 6780E 2P Xeon 6766E 2P Xeon Max 9480 2P Xeon Max 9468 2P Xeon Platinum 8380 2P EPYC 8534P Xeon Platinum 8592+ Xeon 6780E EPYC 8534PN Xeon Platinum 8490H Xeon 6766E Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 2 4 6 8 10 SE +/- 0.003, N = 12 SE +/- 0.005, N = 12 SE +/- 0.010, N = 15 SE +/- 0.002, N = 11 SE +/- 0.004, N = 11 SE +/- 0.003, N = 11 SE +/- 0.008, N = 11 SE +/- 0.006, N = 10 SE +/- 0.008, N = 10 SE +/- 0.012, N = 10 SE +/- 0.005, N = 10 SE +/- 0.004, N = 9 SE +/- 0.003, N = 9 SE +/- 0.006, N = 8 SE +/- 0.007, N = 8 SE +/- 0.003, N = 8 SE +/- 0.003, N = 8 SE +/- 0.019, N = 8 SE +/- 0.003, N = 8 SE +/- 0.010, N = 7 SE +/- 0.006, N = 7 SE +/- 0.006, N = 6 1.199 1.242 1.359 1.921 1.984 2.058 2.187 2.325 2.337 2.605 2.691 3.121 3.366 3.801 4.250 4.324 4.545 4.576 4.806 5.332 6.160 6.673 1. (CXX) g++ options: -O3 CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Primesieve 12.1 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6766E EPYC 9654 EPYC 9684X Xeon 6780E EPYC 9754 2P Xeon Platinum 8380 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9654 2P Xeon 6766E 2P EPYC 9684X 2P Xeon Max 9480 Xeon Max 9468 Xeon 6780E 2P Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 140 280 420 560 700 Min: 16.77 / Avg: 72.3 / Max: 123.74 Min: 16.53 / Avg: 87.74 / Max: 161.22 Min: 21.44 / Avg: 116.27 / Max: 317.57 Min: 43.73 / Avg: 134.25 / Max: 215.68 Min: 26.05 / Avg: 142.51 / Max: 369.88 Min: 33.95 / Avg: 157.31 / Max: 410.23 Min: 43.47 / Avg: 160.81 / Max: 278.45 Min: 39.28 / Avg: 188.39 / Max: 630.34 Min: 52.68 / Avg: 193.17 / Max: 291 Min: 20.57 / Avg: 195.27 / Max: 377.63 Min: 53.72 / Avg: 216.08 / Max: 377.44 Min: 58.4 / Avg: 217.76 / Max: 736.92 Min: 104.53 / Avg: 228.57 / Max: 454.64 Min: 65.08 / Avg: 239.55 / Max: 813.29 Min: 75.05 / Avg: 240.32 / Max: 376.42 Min: 41.72 / Avg: 247.24 / Max: 378.01 Min: 105.81 / Avg: 260.26 / Max: 569.59 Min: 70.6 / Avg: 304.92 / Max: 755.57 Min: 99.97 / Avg: 315.19 / Max: 581.5 Min: 167.18 / Avg: 384.89 / Max: 756.1 Min: 188 / Avg: 419.03 / Max: 755.09 Min: 185.19 / Avg: 432.51 / Max: 754.28
Parallel BZIP2 Compression This test measures the time needed to compress a file (FreeBSD-13.0-RELEASE-amd64-memstick.img) using Parallel BZIP2 compression. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org Seconds, Fewer Is Better Parallel BZIP2 Compression 1.1.13 FreeBSD-13.0-RELEASE-amd64-memstick.img Compression EPYC 9754 2P EPYC 9684X 2P EPYC 9654 2P EPYC 9754 EPYC 9684X EPYC 9654 Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon 6780E 2P Xeon Platinum 8592+ Xeon 6766E 2P Xeon Max 9468 2P Xeon Platinum 8490H Xeon Platinum 8380 2P EPYC 8534P Xeon 6780E Xeon Max 9480 EPYC 8534PN Xeon 6766E Xeon Max 9468 Xeon Platinum 8380 0.7635 1.527 2.2905 3.054 3.8175 SE +/- 0.012053, N = 15 SE +/- 0.015217, N = 15 SE +/- 0.019468, N = 15 SE +/- 0.010826, N = 15 SE +/- 0.015384, N = 15 SE +/- 0.014512, N = 15 SE +/- 0.024402, N = 15 SE +/- 0.017240, N = 15 SE +/- 0.028863, N = 15 SE +/- 0.015943, N = 15 SE +/- 0.023842, N = 15 SE +/- 0.015355, N = 15 SE +/- 0.030907, N = 15 SE +/- 0.007749, N = 10 SE +/- 0.026496, N = 15 SE +/- 0.023387, N = 15 SE +/- 0.014326, N = 10 SE +/- 0.020941, N = 15 SE +/- 0.016947, N = 15 SE +/- 0.012826, N = 10 SE +/- 0.023295, N = 15 SE +/- 0.048155, N = 15 1.068398 1.109446 1.126329 1.355737 1.369063 1.422954 1.432362 1.469210 1.711296 1.718022 1.847537 1.857642 1.982932 2.144857 2.181455 2.200497 2.279971 2.287766 2.321195 2.504621 2.632077 3.393189 1. (CXX) g++ options: -O2 -pthread -lbz2 -lpthread CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better Parallel BZIP2 Compression 1.1.13 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P EPYC 9754 Xeon 6766E Xeon 6780E EPYC 9654 EPYC 9684X Xeon Platinum 8592+ EPYC 9754 2P Xeon Platinum 8380 Xeon 6766E 2P Xeon Platinum 8490H Xeon 6780E 2P EPYC 9654 2P EPYC 9684X 2P Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8592+ 2P Xeon Platinum 8380 2P Xeon Platinum 8490H 2P Xeon Max 9480 2P Xeon Max 9468 2P 120 240 360 480 600 Min: 16.36 / Avg: 50.15 / Max: 132.7 Min: 15.97 / Avg: 55.16 / Max: 153.03 Min: 21.06 / Avg: 71.78 / Max: 246.91 Min: 43.62 / Avg: 72.77 / Max: 199.71 Min: 43.43 / Avg: 79.98 / Max: 238.53 Min: 26.03 / Avg: 87.61 / Max: 297.16 Min: 32.91 / Avg: 100.67 / Max: 337.76 Min: 33.01 / Avg: 124.42 / Max: 372.52 Min: 39.33 / Avg: 126.08 / Max: 368.96 Min: 50.9 / Avg: 141.25 / Max: 290.87 Min: 103.93 / Avg: 141.47 / Max: 301.99 Min: 59.1 / Avg: 149.7 / Max: 374.85 Min: 104.74 / Avg: 150.21 / Max: 342.71 Min: 52.77 / Avg: 151.26 / Max: 452.63 Min: 58 / Avg: 170.11 / Max: 493.47 Min: 73.85 / Avg: 174.26 / Max: 374.82 Min: 72.25 / Avg: 174.54 / Max: 371.49 Min: 71.11 / Avg: 201.02 / Max: 594.06 Min: 95.47 / Avg: 220.44 / Max: 552.38 Min: 162.27 / Avg: 288.67 / Max: 640.02 Min: 193.42 / Avg: 340.67 / Max: 703.88 Min: 199.37 / Avg: 345.8 / Max: 682.95
oneDNN This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.
Result
OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.4 Harness: Deconvolution Batch shapes_3d - Engine: CPU Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P EPYC 9754 2P Xeon Max 9480 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ EPYC 9754 Xeon Max 9468 2P EPYC 9684X 2P EPYC 9654 2P Xeon Platinum 8490H EPYC 9684X EPYC 9654 Xeon Max 9480 Xeon Max 9468 Xeon Platinum 8380 EPYC 8534P EPYC 8534PN Xeon 6780E Xeon 6780E 2P Xeon 6766E Xeon 6766E 2P 0.5619 1.1238 1.6857 2.2476 2.8095 SE +/- 0.000902, N = 9 SE +/- 0.001274, N = 9 SE +/- 0.003127, N = 9 SE +/- 0.001240, N = 9 SE +/- 0.001091, N = 9 SE +/- 0.000392, N = 9 SE +/- 0.001089, N = 9 SE +/- 0.000806, N = 9 SE +/- 0.002070, N = 9 SE +/- 0.001921, N = 9 SE +/- 0.000805, N = 9 SE +/- 0.001272, N = 9 SE +/- 0.001484, N = 9 SE +/- 0.000927, N = 9 SE +/- 0.000508, N = 9 SE +/- 0.006599, N = 9 SE +/- 0.000791, N = 9 SE +/- 0.001685, N = 9 SE +/- 0.002181, N = 9 SE +/- 0.001365, N = 9 SE +/- 0.001097, N = 9 SE +/- 0.001467, N = 9 0.624250 0.736170 0.829653 0.864358 0.867065 0.907161 0.943621 0.945075 1.001914 1.009473 1.077430 1.119230 1.154870 1.196430 1.264070 1.422130 1.498350 1.797490 2.049100 2.245260 2.297120 2.497340 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread CPU Power Consumption
OpenBenchmarking.org Watts, Fewer Is Better oneDNN 3.4 CPU Power Consumption Monitor EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 EPYC 9654 Xeon Platinum 8380 EPYC 9684X Xeon Platinum 8592+ Xeon Platinum 8490H Xeon 6766E 2P Xeon Max 9468 Xeon Max 9480 Xeon 6780E 2P EPYC 9754 2P EPYC 9654 2P EPYC 9684X 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P Xeon Platinum 8490H 2P Xeon Max 9468 2P Xeon Max 9480 2P 130 260 390 520 650 Min: 15.81 / Avg: 62.56 / Max: 121.69 Min: 16.93 / Avg: 70.08 / Max: 140.81 Min: 43.57 / Avg: 98.35 / Max: 172.87 Min: 43.45 / Avg: 112.89 / Max: 208.27 Min: 20.19 / Avg: 122.35 / Max: 259.22 Min: 26.3 / Avg: 135.34 / Max: 274.99 Min: 48.87 / Avg: 151.05 / Max: 290.31 Min: 32.75 / Avg: 155.62 / Max: 314.15 Min: 23.58 / Avg: 173.49 / Max: 373.63 Min: 51.81 / Avg: 187.58 / Max: 359.76 Min: 103.52 / Avg: 196.1 / Max: 321.17 Min: 73.57 / Avg: 201.22 / Max: 367.75 Min: 69.85 / Avg: 202.05 / Max: 376 Min: 105.63 / Avg: 213.96 / Max: 368.1 Min: 40.43 / Avg: 245.22 / Max: 491.94 Min: 53.59 / Avg: 275.56 / Max: 567.73 Min: 62.97 / Avg: 300.29 / Max: 621 Min: 98.61 / Avg: 304.22 / Max: 566.32 Min: 70.82 / Avg: 355.72 / Max: 740.88 Min: 169.45 / Avg: 409.27 / Max: 716.88 Min: 194.92 / Avg: 425.54 / Max: 706.02 Min: 187.54 / Avg: 431.59 / Max: 733.22
CPU Power Consumption Monitor OpenBenchmarking.org Watts CPU Power Consumption Monitor Phoronix Test Suite System Monitoring EPYC 8534PN EPYC 8534P Xeon 6766E Xeon 6780E EPYC 9754 Xeon Platinum 8380 EPYC 9654 Xeon Platinum 8592+ Xeon Platinum 8490H EPYC 9684X Xeon 6766E 2P Xeon Max 9468 Xeon Max 9480 Xeon 6780E 2P EPYC 9754 2P Xeon Platinum 8380 2P Xeon Platinum 8592+ 2P EPYC 9654 2P Xeon Platinum 8490H 2P EPYC 9684X 2P Xeon Max 9468 2P Xeon Max 9480 2P 170 340 510 680 850 Min: 9.26 / Avg: 114.13 / Max: 155.71 Min: 8.43 / Avg: 128.7 / Max: 184.62 Min: 21.86 / Avg: 142.1 / Max: 271.22 Min: 21.7 / Avg: 169.1 / Max: 357.69 Min: 10.09 / Avg: 225.05 / Max: 397.44 Min: 26.31 / Avg: 237.91 / Max: 292.14 Min: 14.69 / Avg: 244.69 / Max: 373.1 Min: 17.77 / Avg: 260.08 / Max: 405.42 Min: 30.38 / Avg: 278.1 / Max: 383.24 Min: 17.06 / Avg: 290.08 / Max: 413.51 Min: 52.85 / Avg: 291.05 / Max: 542.48 Min: 38.11 / Avg: 291.53 / Max: 379.76 Min: 38.16 / Avg: 295.53 / Max: 384.53 Min: 53.44 / Avg: 321.01 / Max: 713.33 Min: 20.22 / Avg: 375.51 / Max: 735.74 Min: 51.38 / Avg: 437.37 / Max: 583.47 Min: 39.72 / Avg: 455.07 / Max: 946.96 Min: 35.7 / Avg: 464.49 / Max: 751.79 Min: 87.48 / Avg: 517.33 / Max: 758.17 Min: 33.66 / Avg: 525.49 / Max: 837.69 Min: 102.49 / Avg: 566.6 / Max: 756.48 Min: 95.52 / Avg: 569.8 / Max: 760.61
Xeon Platinum 8490H Processor: Intel Xeon Platinum 8490H @ 3.50GHz (60 Cores / 120 Threads), Motherboard: Quanta Cloud QuantaGrid D54Q-2U S6Q-MB-MPS (3B05.TEL4P1 BIOS), Chipset: Intel Device 1bce, Memory: 512GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b000590Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 16 April 2024 00:42 by user phoronix.
Xeon Platinum 8490H 2P Processor: 2 x Intel Xeon Platinum 8490H @ 3.50GHz (120 Cores / 240 Threads), Motherboard: Quanta Cloud QuantaGrid D54Q-2U S6Q-MB-MPS (3B05.TEL4P1 BIOS), Chipset: Intel Device 1bce, Memory: 1008GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: 2 x Intel X710 for 10GBASE-T
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b000590Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 17 April 2024 09:38 by user phoronix.
Xeon Platinum 8592+ Processor: INTEL XEON PLATINUM 8592+ @ 3.90GHz (64 Cores / 128 Threads), Motherboard: Quanta Cloud QuantaGrid D54Q-2U S6Q-MB-MPS (3B05.TEL4P1 BIOS), Chipset: Intel Device 1bce, Memory: 512GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000200Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 11 April 2024 18:15 by user phoronix.
Xeon Platinum 8592+ 2P Processor: 2 x INTEL XEON PLATINUM 8592+ @ 3.90GHz (128 Cores / 256 Threads), Motherboard: Quanta Cloud QuantaGrid D54Q-2U S6Q-MB-MPS (3B05.TEL4P1 BIOS), Chipset: Intel Device 1bce, Memory: 1008GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: 2 x Intel X710 for 10GBASE-T
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000200Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 8 April 2024 10:49 by user phoronix.
Xeon Max 9468 Processor: Intel Xeon Max 9468 @ 3.50GHz (48 Cores / 96 Threads), Motherboard: Quanta Cloud QuantaGrid D54Q-2U S6Q-MB-MPS (3B05.TEL4P1 BIOS), Chipset: Intel Device 1bce, Memory: 576GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2c000290Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 21 April 2024 09:58 by user phoronix.
Xeon Max 9468 2P Processor: 2 x Intel Xeon Max 9468 @ 3.50GHz (96 Cores / 192 Threads), Motherboard: Quanta Cloud QuantaGrid D54Q-2U S6Q-MB-MPS (3B05.TEL4P1 BIOS), Chipset: Intel Device 1bce, Memory: 1136GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: 2 x Intel X710 for 10GBASE-T
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2c000290Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 22 April 2024 15:30 by user phoronix.
Xeon Max 9480 Processor: Intel Xeon Max 9480 @ 3.50GHz (56 Cores / 112 Threads), Motherboard: Quanta Cloud QuantaGrid D54Q-2U S6Q-MB-MPS (3B05.TEL4P1 BIOS), Chipset: Intel Device 1bce, Memory: 576GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2c000290Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 26 April 2024 09:38 by user phoronix.
Xeon Max 9480 2P Processor: 2 x Intel Xeon Max 9480 @ 3.50GHz (112 Cores / 224 Threads), Motherboard: Quanta Cloud QuantaGrid D54Q-2U S6Q-MB-MPS (3B05.TEL4P1 BIOS), Chipset: Intel Device 1bce, Memory: 1136GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: 2 x Intel X710 for 10GBASE-T
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2c000290Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 27 April 2024 16:18 by user phoronix.
Xeon Platinum 8380 2P Processor: 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads), Motherboard: Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS), Chipset: Intel Ice Lake IEH, Memory: 512GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Monitor: VE228, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1080
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd0003d1Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Mitigation of Microcode + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 7 May 2024 16:33 by user phoronix.
Xeon Platinum 8380 Processor: Intel Xeon Platinum 8380 @ 3.40GHz (40 Cores / 80 Threads), Motherboard: Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS), Chipset: Intel Ice Lake IEH, Memory: 256GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Monitor: VE228, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1080
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd0003d1Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Mitigation of Microcode + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 9 May 2024 13:53 by user phoronix.
EPYC 9684X 2P Processor: 2 x AMD EPYC 9684X 96-Core @ 2.55GHz (192 Cores / 384 Threads), Motherboard: AMD Titanite_4G (RTI1007B BIOS), Chipset: AMD Device 14a4, Memory: 1520GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: Broadcom NetXtreme BCM5720 PCIe
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa101144Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 11 May 2024 02:04 by user phoronix.
EPYC 9684X Processor: AMD EPYC 9684X 96-Core @ 2.55GHz (96 Cores / 192 Threads), Motherboard: AMD Titanite_4G (RTI1007B BIOS), Chipset: AMD Device 14a4, Memory: 12 x 64GB DDR5-4800MT/s Samsung M321R8GA0BB0-CQKEG, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: Broadcom NetXtreme BCM5720 PCIe
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa101144Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 12 May 2024 12:29 by user phoronix.
EPYC 9654 Processor: AMD EPYC 9654 96-Core @ 2.40GHz (96 Cores / 192 Threads), Motherboard: AMD Titanite_4G (RTI1007B BIOS), Chipset: AMD Device 14a4, Memory: 12 x 64GB DDR5-4800MT/s Samsung M321R8GA0BB0-CQKEG, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: Broadcom NetXtreme BCM5720 PCIe
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa101144Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 13 May 2024 14:14 by user phoronix.
EPYC 9654 2P Processor: 2 x AMD EPYC 9654 96-Core @ 2.40GHz (192 Cores / 384 Threads), Motherboard: AMD Titanite_4G (RTI1007B BIOS), Chipset: AMD Device 14a4, Memory: 1520GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: Broadcom NetXtreme BCM5720 PCIe
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa101144Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 14 May 2024 15:08 by user phoronix.
EPYC 9754 Processor: AMD EPYC 9754 128-Core @ 2.25GHz (128 Cores / 256 Threads), Motherboard: AMD Titanite_4G (RTI1007B BIOS), Chipset: AMD Device 14a4, Memory: 12 x 64GB DDR5-4800MT/s Samsung M321R8GA0BB0-CQKEG, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: Broadcom NetXtreme BCM5720 PCIe
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa00116Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 15 May 2024 21:09 by user phoronix.
EPYC 9754 2P Processor: 2 x AMD EPYC 9754 128-Core @ 2.25GHz (256 Cores / 512 Threads), Motherboard: AMD Titanite_4G (RTI1007B BIOS), Chipset: AMD Device 14a4, Memory: 1520GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: Broadcom NetXtreme BCM5720 PCIe
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa00116Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 17 May 2024 00:34 by user phoronix.
EPYC 8534P Processor: AMD EPYC 8534P 64-Core @ 2.30GHz (64 Cores / 128 Threads), Motherboard: AMD Cinnabar (RCB1009C BIOS), Chipset: AMD Device 14a4, Memory: 6 x 32GB DDR5-4800MT/s Samsung M321R4GA0BB0-CQKMG, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92 + 2000GB Corsair MP700 PRO, Graphics: ASPEED, Network: 2 x Broadcom NetXtreme BCM5720 PCIe
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa00213Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 18 May 2024 13:07 by user phoronix.
EPYC 8534PN Processor: AMD EPYC 8534PN 64-Core @ 2.00GHz (64 Cores / 128 Threads), Motherboard: AMD Cinnabar (RCB1009C BIOS), Chipset: AMD Device 14a4, Memory: 6 x 32GB DDR5-4800MT/s Samsung M321R4GA0BB0-CQKMG, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: 2 x Broadcom NetXtreme BCM5720 PCIe
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa00213Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 19 May 2024 17:06 by user phoronix.
Xeon 6766E 2P Processor: 2 x Intel Xeon 6766E @ 2.70GHz (288 Cores), Motherboard: Quanta Cloud QuantaGrid D55Q-2U S7Q-MB-MPS-MDP (2A05.QCT001 BIOS), Chipset: Intel Ice Lake IEH, Memory: 512GB, Disk: 3201GB Micron_7450_MTFDKCB3T2TFS + 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: 2 x Broadcom BCM57416 NetXtreme-E Dual-Media 10G RDMA
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x13000131Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 24 May 2024 12:45 by user phoronix.
Xeon 6766E Processor: Intel Xeon 6766E @ 2.70GHz (144 Cores), Motherboard: Quanta Cloud QuantaGrid D55Q-2U S7Q-MB-MPS-MDP (2A05.QCT001 BIOS), Chipset: Intel Ice Lake IEH, Memory: 256GB, Disk: 2 x 1920GB KIOXIA KCD8XPUG1T92 + 3201GB Micron_7450_MTFDKCB3T2TFS, Graphics: ASPEED, Network: 2 x Broadcom BCM57416 NetXtreme-E Dual-Media 10G RDMA
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x13000131Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 25 May 2024 14:43 by user phoronix.
Xeon 6780E Processor: Intel Xeon 6780E @ 3.00GHz (144 Cores), Motherboard: Quanta Cloud QuantaGrid D55Q-2U S7Q-MB-MPS-MDP (2A05.QCT001 BIOS), Chipset: Intel Ice Lake IEH, Memory: 256GB, Disk: 3201GB Micron_7450_MTFDKCB3T2TFS + 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: 2 x Broadcom BCM57416 NetXtreme-E Dual-Media 10G RDMA
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x13000131Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 26 May 2024 18:38 by user phoronix.
Xeon 6780E 2P Processor: 2 x Intel Xeon 6780E @ 3.00GHz (288 Cores), Motherboard: Quanta Cloud QuantaGrid D55Q-2U S7Q-MB-MPS-MDP (2A05.QCT001 BIOS), Chipset: Intel Ice Lake IEH, Memory: 512GB, Disk: 3201GB Micron_7450_MTFDKCB3T2TFS + 2 x 1920GB KIOXIA KCD8XPUG1T92, Graphics: ASPEED, Network: 2 x Broadcom BCM57416 NetXtreme-E Dual-Media 10G RDMA
OS: Ubuntu 24.04, Kernel: 6.9.0-060900rc3-generic (x86_64), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x13000131Java Notes: OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Notes: Python 3.12.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 27 May 2024 18:06 by user phoronix.