Sapphire Rapids AVX-512 Benchmarks

Ice Lake, Sapphire Rapids, Genoa AVX-512 benchmarks by Michael Larabel for a future article.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2301176-NE-SAPPHIRER14
Jump To Table - Results

View

Do Not Show Noisy Results
Do Not Show Results With Incomplete Data
Do Not Show Results With Little Change/Spread
List Notable Results

Limit displaying results to tests within:

CPU Massive 3 Tests
Creator Workloads 7 Tests
Cryptography 2 Tests
Game Development 2 Tests
HPC - High Performance Computing 7 Tests
Machine Learning 5 Tests
Multi-Core 8 Tests
Intel oneAPI 7 Tests
Python Tests 2 Tests
Raytracing 2 Tests
Renderers 2 Tests
Server CPU Tests 2 Tests

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Geometric Means Per-Suite/Category
Show Wins / Losses Counts (Pie Chart)
Normalize Results
Remove Outliers Before Calculating Averages

Graph Settings

Force Line Graphs Where Applicable
Convert To Scalar Where Applicable
Disable Color Branding
Prefer Vertical Bar Graphs
No Box Plots
On Line Graphs With Missing Data, Connect The Line Gaps

Multi-Way Comparison

Condense Comparison
Transpose Comparison

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
Xeon 8380: AVX-512 Off
January 13 2023
  9 Hours, 51 Minutes
Xeon 8380: AVX-512 On
January 13 2023
  10 Hours, 13 Minutes
Xeon 8490H: AVX-512 Off
January 12 2023
  7 Hours, 13 Minutes
Xeon 8490H: AVX-512 On
January 11 2023
  7 Hours, 3 Minutes
EPYC 9654: AVX-512 Off
January 15 2023
  7 Hours, 12 Minutes
EPYC 9654: AVX-512 On
January 16 2023
  8 Hours, 47 Minutes
Invert Hiding All Results Option
  8 Hours, 23 Minutes

Only show results where is faster than
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):

Xeon 8380: AVX-512 Off

Kernel Notes: Transparent Huge Pages: madvise
Environment Notes: CXXFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile" CFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile"
Compiler 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-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-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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 -v
Processor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd000375
Python Notes: Python 3.10.7
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + retbleed: 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 IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected

Xeon 8380: AVX-512 On

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 SAMSUNG MZWLJ1T9HBJR-00007, Graphics: ASPEED, Monitor: VGA HDMI, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP

OS: Ubuntu 22.10, Kernel: 5.19.0-29-generic (x86_64), Desktop: GNOME Shell 43.1, Display Server: X Server 1.21.1.4, Vulkan: 1.3.224, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1920x1080

Kernel Notes: Transparent Huge Pages: madvise
Environment Notes: CXXFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile" CFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile"
Compiler 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-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-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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 -v
Processor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd000375
Python Notes: Python 3.10.7
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + retbleed: 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 IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected

Xeon 8490H: AVX-512 Off

Kernel Notes: Transparent Huge Pages: madvise
Environment Notes: CXXFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile" CFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile"
Compiler 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-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-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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 -v
Processor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b0000c0
Python Notes: Python 3.10.7
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected

Xeon 8490H: AVX-512 On

Processor: 2 x Intel Xeon Platinum 8490H @ 3.50GHz (120 Cores / 240 Threads), Motherboard: Quanta Cloud S6Q-MB-MPS (3A10.uh BIOS), Chipset: Intel Device 1bce, Memory: 1008GB, Disk: 2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007, Graphics: ASPEED, Monitor: VGA HDMI, Network: 4 x Intel E810-C for QSFP + 2 x Intel X710 for 10GBASE-T

Kernel Notes: Transparent Huge Pages: madvise
Environment Notes: CXXFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile" CFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile"
Compiler 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-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-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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 -v
Processor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b0000c0
Python Notes: Python 3.10.7
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected

EPYC 9654: AVX-512 Off

Kernel Notes: Transparent Huge Pages: madvise
Environment Notes: CXXFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile" CFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile"
Compiler 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-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-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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 -v
Processor Notes: Scaling Governor: amd-pstate performance (Boost: Enabled) - CPU Microcode: 0xa10110d
Python Notes: Python 3.10.7
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

EPYC 9654: AVX-512 On

Processor: 2 x AMD EPYC 9654 96-Core @ 3.71GHz (192 Cores / 384 Threads), Motherboard: AMD Titanite_4G (RTI1002E BIOS), Chipset: AMD Device 14a4, Memory: 1520GB, Disk: 2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007, Graphics: ASPEED, Monitor: VGA HDMI, Network: Broadcom NetXtreme BCM5720 PCIe

Kernel Notes: Transparent Huge Pages: madvise
Environment Notes: CXXFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile" CFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile"
Compiler 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-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-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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 -v
Processor Notes: Scaling Governor: amd-pstate performance (Boost: Enabled) - CPU Microcode: 0xa10110d
Python Notes: Python 3.10.7
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

ProcessorMotherboardChipsetMemoryDiskGraphicsMonitorNetworkOSKernelDesktopDisplay ServerVulkanCompilerFile-SystemScreen ResolutionXeon 8380Xeon 8490HEPYC 9654 AVX-512 Off AVX-512 On AVX-512 Off AVX-512 On AVX-512 Off AVX-512 On2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads)Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS)Intel Ice Lake IEH512GB2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007ASPEEDVGA HDMI2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFPUbuntu 22.105.19.0-29-generic (x86_64)GNOME Shell 43.1X Server 1.21.1.41.3.224GCC 12.2.0ext41920x10802 x Intel Xeon Platinum 8490H @ 3.50GHz (120 Cores / 240 Threads)Quanta Cloud S6Q-MB-MPS (3A10.uh BIOS)Intel Device 1bce1008GB4 x Intel E810-C for QSFP + 2 x Intel X710 for 10GBASE-T2 x AMD EPYC 9654 96-Core @ 3.71GHz (192 Cores / 384 Threads)AMD Titanite_4G (RTI1002E BIOS)AMD Device 14a41520GBBroadcom NetXtreme BCM5720 PCIeOpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseEnvironment Details- Xeon 8380: AVX-512 Off: CXXFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile" CFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile"- Xeon 8380: AVX-512 On: CXXFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile" CFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile" - Xeon 8490H: AVX-512 Off: CXXFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile" CFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile"- Xeon 8490H: AVX-512 On: CXXFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile" CFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile" - EPYC 9654: AVX-512 Off: CXXFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile" CFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile"- EPYC 9654: AVX-512 On: CXXFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile" CFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile" Compiler Details- --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-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-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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 -v Processor Details- Xeon 8380: AVX-512 Off: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd000375- Xeon 8380: AVX-512 On: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd000375- Xeon 8490H: AVX-512 Off: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b0000c0- Xeon 8490H: AVX-512 On: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b0000c0- EPYC 9654: AVX-512 Off: Scaling Governor: amd-pstate performance (Boost: Enabled) - CPU Microcode: 0xa10110d- EPYC 9654: AVX-512 On: Scaling Governor: amd-pstate performance (Boost: Enabled) - CPU Microcode: 0xa10110dPython Details- Python 3.10.7Security Details- Xeon 8380: AVX-512 Off: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + retbleed: 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 IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected - Xeon 8380: AVX-512 On: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + retbleed: 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 IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected - Xeon 8490H: AVX-512 Off: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected - Xeon 8490H: AVX-512 On: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9654: AVX-512 Off: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9654: AVX-512 On: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

openvino: Vehicle Detection FP16 - CPUopenvino: Vehicle Detection FP16-INT8 - CPUcpuminer-opt: LBC, LBRY Creditscpuminer-opt: Quad SHA-256, Pyriteopenvino: Vehicle Detection FP16-INT8 - CPUcpuminer-opt: Triple SHA-256, Onecoincpuminer-opt: Skeincoindeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Streamminibude: OpenMP - BM2minibude: OpenMP - BM2cpuminer-opt: scryptcpuminer-opt: x25xminibude: OpenMP - BM1minibude: OpenMP - BM1deepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Streammnn: squeezenetv1.1deepsparse: CV Detection,YOLOv5s COCO - Asynchronous Multi-Streamopenfoam: drivaerFastback, Medium Mesh Size - Execution Timedeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamospray: gravity_spheres_volume/dim_512/scivis/real_timeospray: gravity_spheres_volume/dim_512/ao/real_timeopenvino: Vehicle Detection FP16 - CPUmnn: resnet-v2-50embree: Pathtracer ISPC - Crowncpuminer-opt: Deepcoinospray-studio: 3 - 4K - 32 - Path Tracerdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamonednn: Deconvolution Batch shapes_3d - f32 - CPUospray: gravity_spheres_volume/dim_512/pathtracer/real_timeospray-studio: 1 - 4K - 32 - Path Tracerospray-studio: 1 - 4K - 16 - Path Tracerospray-studio: 3 - 4K - 16 - Path Tracerospray-studio: 1 - 4K - 1 - Path Tracerospray-studio: 3 - 4K - 1 - Path Traceropenfoam: drivaerFastback, Large Mesh Size - Execution Timedeepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Streamembree: Pathtracer ISPC - Asian Dragonlczero: Eigenopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUmnn: SqueezeNetV1.0mnn: mobilenetV3lczero: BLASdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Synchronous Single-Streamdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Synchronous Single-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Streamoidn: RT.hdr_alb_nrm.3840x2160oidn: RTLightmap.hdr.4096x4096deepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: CV Detection,YOLOv5s COCO - Synchronous Single-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamopenvkl: vklBenchmark ISPCopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUdeepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Streamdeepsparse: CV Detection,YOLOv5s COCO - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamsmhasher: FarmHash32 x86_64 AVXopenfoam: drivaerFastback, Large Mesh Size - Mesh Timeopenfoam: drivaerFastback, Medium Mesh Size - Mesh Timesmhasher: FarmHash32 x86_64 AVXdeepsparse: CV Detection,YOLOv5s COCO - Synchronous Single-StreamXeon 8380Xeon 8490HEPYC 9654 AVX-512 Off AVX-512 On AVX-512 Off AVX-512 On AVX-512 Off AVX-512 On1014.2711.411570833551731750.30647033399350158.52611993.06779.7231122.741840.4171.8461796.159541.8617183.83252.678247.2732358.63124379.579213.779214.345519.6818.13370.5079579306278043.114043.16011.9928619.9578481652394328490149117707106.76118.8158113.326196.3441470950413.984.1951.764478045.854721.797359.515916.799659.417516.82752.071.03251.567873.746369.7264913.4196916.93607921.58217.0993161.3342105.262430330.53887.20762162.7230631.58414.32861119.454.494191339240404442.951328547613273221.72942588.758103.5502307.422650.3693.5732339.323835.7846227.45442.297318.7928361.36688455.031221.064021.224217.837.90974.8033646904984748.413448.23930.87519625.5179418342087224927130815637117.00594.3750228.1219106.1798575067701.603.9991.709552479.140512.626449.420820.231349.514520.19332.941.42180.170447.821683.4197821.9588819.53259171.16175.6914125.163987.865833979.991051.3525160.664531.08411.97691732.7430.912639176873203878.791092637677733232.20852602.652104.1061623.882166.9099.2252480.607832.8542264.77043.551405.7230194.78975584.148722.832823.174817.2917.71662.0377911073730066.754066.72812.0124132.483731373156011860898011704319.12455.3242187.647790.3959824373906.765.2222.288754246.861421.321247.209421.178946.973521.28622.951.48257.562271.934684.8378889.3899887.589310341.57226.2678147.4578102.613637809.711039.3153169.8416423.26811.78233738.9219.3167267714726206206.7220742001072453355.66863643.840145.7543665.554071.59143.4513586.2731492.8844424.27142.737592.4522195.47713875.641635.954436.36438.018.49692.65121055133274992.021192.03280.73682242.966227204135891633285910294305.12423.5509281.2361122.526011134103184.014.4152.0551067183.700811.937837.247326.843236.938227.06704.191.99168.228540.1407100.5397647.7545648.199511961.04141.1445101.042568.434742508.35847.91029159.6732923.2739.97343791.807.7149278013343636218.2533388501413063491.34266371.320254.8532947.945712.58207.6975192.4251410.4222510.87757.566660.5371110.949481006.082425.309626.395312.6424.081153.181516524026541120.5142120.51891.7366740.66582247511216133417058362803.54444.8444206.3171177.67748161114083.249.2783.889942571.508813.974032.826330.456332.402430.85533.461.72194.717967.9016109.2581787.7235787.327011761.14187.3515144.885895.166536458.31867.379146.7545326.4139.18357420.494.341068993225250011029.4533262871990790704.58188562.694342.5084816.317691.78290.9627274.0501966.4052613.93557.204809.0751112.381691205.181343.235843.70816.4614.727181.88891602902212084.867684.39960.98621454.1553185009260110585816952849.6115.2240191.3325213.95129085118103.948.7623.8669171100.28559.966728.977134.500829.202934.23223.481.66135.993748.7003126.01141132.61331128.845013410.96155.9118118.352579.496739795.36996.77741145.8994426.5517.9604OpenBenchmarking.org

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16 - Device: CPUAVX-512 OffAVX-512 On16003200480064008000SE +/- 0.71, N = 3SE +/- 0.97, N = 3SE +/- 0.62, N = 3SE +/- 1.23, N = 3SE +/- 2.86, N = 3SE +/- 1.27, N = 31014.271119.451732.743738.923791.807420.491. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fsigned-char -ffunction-sections -fdata-sections -fno-strict-overflow -fwrapv
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16 - Device: CPUAVX-512 OffAVX-512 On13002600390052006500Min: 1013.52 / Avg: 1014.27 / Max: 1015.69Min: 1117.86 / Avg: 1119.45 / Max: 1121.22Min: 1731.65 / Avg: 1732.74 / Max: 1733.81Min: 3736.64 / Avg: 3738.92 / Max: 3740.84Min: 3788.73 / Avg: 3791.8 / Max: 3797.51Min: 7417.96 / Avg: 7420.49 / Max: 7421.971. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fsigned-char -ffunction-sections -fdata-sections -fno-strict-overflow -fwrapv

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16-INT8 - Device: CPUAVX-512 OffAVX-512 On714212835SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 311.414.4930.9119.317.714.341. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fsigned-char -ffunction-sections -fdata-sections -fno-strict-overflow -fwrapv
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2022.3CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2952.33 / Max: 3400Min: 2300 / Avg: 2464.76 / Max: 3400Min: 2327 / Avg: 2504.87 / Max: 3513Min: 1900 / Avg: 2571.73 / Max: 3502Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 124.05 / Avg: 517.82 / Max: 584.06Min: 127.49 / Avg: 518.52 / Max: 595.86Min: 209.15 / Avg: 668.46 / Max: 770.95Min: 211.57 / Avg: 667.99 / Max: 782.09Min: 57.68 / Avg: 622.09 / Max: 684.99Min: 57.77 / Avg: 594.48 / Max: 654.28
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOpenVINO 2022.3CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 54 / Avg: 74.99 / Max: 79Min: 53 / Avg: 74.24 / Max: 79Min: 40 / Avg: 53.23 / Max: 57Min: 40 / Avg: 52.32 / Max: 55Min: 37.25 / Avg: 52.53 / Max: 56Min: 37.75 / Avg: 51.65 / Max: 54.25
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16-INT8 - Device: CPUAVX-512 OffAVX-512 On714212835Min: 11.38 / Avg: 11.41 / Max: 11.44Min: 4.48 / Avg: 4.49 / Max: 4.49Min: 30.9 / Avg: 30.91 / Max: 30.93Min: 19.3 / Avg: 19.31 / Max: 19.32Min: 7.71 / Avg: 7.71 / Max: 7.71Min: 4.34 / Avg: 4.34 / Max: 4.341. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fsigned-char -ffunction-sections -fdata-sections -fno-strict-overflow -fwrapv

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: LBC, LBRY CreditsAVX-512 OffAVX-512 On200K400K600K800K1000KSE +/- 498.81, N = 3SE +/- 390.61, N = 3SE +/- 820.11, N = 3SE +/- 1047.73, N = 3SE +/- 1744.51, N = 3SE +/- 3825.46, N = 315708341913326391767267749278010689931. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s Per Watt, More Is BetterCpuminer-Opt 3.20.3Algorithm: LBC, LBRY CreditsAVX-512 OffAVX-512 On400800120016002000314.92841.20404.761033.12834.911639.75
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterCpuminer-Opt 3.20.3CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2737.13 / Max: 3400Min: 2300 / Avg: 2623.57 / Max: 3400Min: 1900 / Avg: 2760.78 / Max: 3502Min: 1900 / Avg: 2817.16 / Max: 3501Min: 2733 / Avg: 3270.61 / Max: 3709Min: 3092 / Avg: 3394.73 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterCpuminer-Opt 3.20.3CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 119.65 / Avg: 498.8 / Max: 602.02Min: 123.8 / Avg: 498.26 / Max: 601.23Min: 132.94 / Avg: 652.03 / Max: 781.58Min: 205.28 / Avg: 651.11 / Max: 781.83Min: 58.49 / Avg: 590.22 / Max: 704.35Min: 31.53 / Avg: 651.92 / Max: 765.29
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterCpuminer-Opt 3.20.3CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 52 / Avg: 69.81 / Max: 74Min: 52 / Avg: 70.74 / Max: 76Min: 40 / Avg: 54.02 / Max: 57Min: 39 / Avg: 52.42 / Max: 56Min: 36.75 / Avg: 51.26 / Max: 55.13Min: 37.25 / Avg: 53.37 / Max: 57.25
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: LBC, LBRY CreditsAVX-512 OffAVX-512 On200K400K600K800K1000KMin: 156360 / Avg: 157083.33 / Max: 158040Min: 418620 / Avg: 419133.33 / Max: 419900Min: 262970 / Avg: 263916.67 / Max: 265550Min: 671440 / Avg: 672676.67 / Max: 674760Min: 490730 / Avg: 492780 / Max: 496250Min: 1061350 / Avg: 1068993.33 / Max: 10731101. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: Quad SHA-256, PyriteAVX-512 OffAVX-512 On500K1000K1500K2000K2500KSE +/- 4107.35, N = 3SE +/- 1289.92, N = 3SE +/- 657.95, N = 3SE +/- 3990.27, N = 3SE +/- 7009.29, N = 3SE +/- 10844.37, N = 33551739240406873201472620133436322525001. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s Per Watt, More Is BetterCpuminer-Opt 3.20.3Algorithm: Quad SHA-256, PyriteAVX-512 OffAVX-512 On8001600240032004000710.081859.261053.062276.982351.123555.70
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterCpuminer-Opt 3.20.3CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2868.13 / Max: 3101Min: 2300 / Avg: 2651.15 / Max: 3400Min: 1900 / Avg: 2927.09 / Max: 3501Min: 1900 / Avg: 2857.47 / Max: 3507Min: 3033 / Avg: 3343.09 / Max: 3709Min: 3114 / Avg: 3410.31 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterCpuminer-Opt 3.20.3CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 120.7 / Avg: 500.19 / Max: 602.54Min: 79.08 / Avg: 496.99 / Max: 602.19Min: 220 / Avg: 652.69 / Max: 740.32Min: 206.88 / Avg: 646.74 / Max: 779.66Min: 54.9 / Avg: 567.54 / Max: 675.79Min: 32.24 / Avg: 633.49 / Max: 739.43
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterCpuminer-Opt 3.20.3CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 51 / Avg: 70.21 / Max: 75Min: 52 / Avg: 70.84 / Max: 75Min: 39 / Avg: 53.91 / Max: 58Min: 39 / Avg: 52.72 / Max: 57Min: 36.38 / Avg: 50.97 / Max: 54.25Min: 38.13 / Avg: 52.94 / Max: 56.13
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: Quad SHA-256, PyriteAVX-512 OffAVX-512 On400K800K1200K1600K2000KMin: 346960 / Avg: 355173.33 / Max: 359410Min: 922560 / Avg: 924040 / Max: 926610Min: 686450 / Avg: 687320 / Max: 688610Min: 1468550 / Avg: 1472620 / Max: 1480600Min: 1320850 / Avg: 1334363.33 / Max: 1344350Min: 2232080 / Avg: 2252500 / Max: 22690401. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16-INT8 - Device: CPUAVX-512 OffAVX-512 On2K4K6K8K10KSE +/- 2.52, N = 3SE +/- 2.81, N = 3SE +/- 1.04, N = 3SE +/- 1.24, N = 3SE +/- 0.29, N = 3SE +/- 1.97, N = 31750.304442.953878.796206.726218.2511029.451. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fsigned-char -ffunction-sections -fdata-sections -fno-strict-overflow -fwrapv
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16-INT8 - Device: CPUAVX-512 OffAVX-512 On2K4K6K8K10KMin: 1746.05 / Avg: 1750.3 / Max: 1754.78Min: 4438.93 / Avg: 4442.95 / Max: 4448.37Min: 3876.78 / Avg: 3878.79 / Max: 3880.22Min: 6204.35 / Avg: 6206.72 / Max: 6208.56Min: 6217.68 / Avg: 6218.25 / Max: 6218.57Min: 11026.7 / Avg: 11029.45 / Max: 11033.271. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fsigned-char -ffunction-sections -fdata-sections -fno-strict-overflow -fwrapv

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: Triple SHA-256, OnecoinAVX-512 OffAVX-512 On700K1400K2100K2800K3500KSE +/- 579.55, N = 3SE +/- 8466.21, N = 3SE +/- 3398.36, N = 3SE +/- 18201.57, N = 3SE +/- 18173.37, N = 3SE +/- 23073.29, N = 15647033132854710926372074200333885033262871. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s Per Watt, More Is BetterCpuminer-Opt 3.20.3Algorithm: Triple SHA-256, OnecoinAVX-512 OffAVX-512 On120024003600480060001301.102672.901741.893203.655632.974975.70
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterCpuminer-Opt 3.20.3CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2888.78 / Max: 3283Min: 2300 / Avg: 2643.21 / Max: 3400Min: 1900 / Avg: 2957.05 / Max: 3510Min: 1900 / Avg: 2875.21 / Max: 3511Min: 3008 / Avg: 3335.89 / Max: 3709Min: 3178 / Avg: 3465.81 / Max: 3714
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterCpuminer-Opt 3.20.3CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 75.22 / Avg: 497.3 / Max: 601.9Min: 119.7 / Avg: 497.04 / Max: 601.36Min: 221.38 / Avg: 627.27 / Max: 771.94Min: 206.27 / Avg: 647.45 / Max: 782.3Min: 57.41 / Avg: 592.73 / Max: 701.33Min: 33.08 / Avg: 668.51 / Max: 793.29
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterCpuminer-Opt 3.20.3CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 51 / Avg: 69.9 / Max: 74Min: 51 / Avg: 70.07 / Max: 75Min: 40 / Avg: 51.93 / Max: 56Min: 39 / Avg: 52.28 / Max: 56Min: 35.88 / Avg: 50.93 / Max: 54.25Min: 36.75 / Avg: 55.17 / Max: 58.75
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: Triple SHA-256, OnecoinAVX-512 OffAVX-512 On600K1200K1800K2400K3000KMin: 646390 / Avg: 647033.33 / Max: 648190Min: 1317220 / Avg: 1328546.67 / Max: 1345110Min: 1085840 / Avg: 1092636.67 / Max: 1096060Min: 2052090 / Avg: 2074200 / Max: 2110300Min: 3318430 / Avg: 3338850 / Max: 3375100Min: 3248420 / Avg: 3326286.67 / Max: 35094801. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: SkeincoinAVX-512 OffAVX-512 On400K800K1200K1600K2000KSE +/- 716.54, N = 3SE +/- 438.80, N = 3SE +/- 329.36, N = 3SE +/- 668.14, N = 3SE +/- 7503.92, N = 3SE +/- 4597.66, N = 33993506132736777331072453141306319907901. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s Per Watt, More Is BetterCpuminer-Opt 3.20.3Algorithm: SkeincoinAVX-512 OffAVX-512 On7001400210028003500800.441225.351035.241658.462428.443141.21
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterCpuminer-Opt 3.20.3CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2776.23 / Max: 3195Min: 2300 / Avg: 2673.53 / Max: 3063Min: 1900 / Avg: 2872.56 / Max: 3501Min: 2700 / Avg: 2861.18 / Max: 3505Min: 2896 / Avg: 3299.49 / Max: 3709Min: 3001 / Avg: 3367.28 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterCpuminer-Opt 3.20.3CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 118.62 / Avg: 498.92 / Max: 601.83Min: 123.73 / Avg: 500.49 / Max: 601.47Min: 224.85 / Avg: 654.66 / Max: 780.76Min: 103.7 / Avg: 646.66 / Max: 782.6Min: 53.17 / Avg: 581.88 / Max: 686.77Min: 34.28 / Avg: 633.77 / Max: 745.91
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterCpuminer-Opt 3.20.3CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 51 / Avg: 69.73 / Max: 74Min: 52 / Avg: 71.11 / Max: 77Min: 40 / Avg: 53.91 / Max: 58Min: 39 / Avg: 52.76 / Max: 56Min: 36.38 / Avg: 50.92 / Max: 54.25Min: 36.38 / Avg: 52.6 / Max: 56.13
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: SkeincoinAVX-512 OffAVX-512 On300K600K900K1200K1500KMin: 398140 / Avg: 399350 / Max: 400620Min: 612680 / Avg: 613273.33 / Max: 614130Min: 677360 / Avg: 677733.33 / Max: 678390Min: 1071340 / Avg: 1072453.33 / Max: 1073650Min: 1404250 / Avg: 1413063.33 / Max: 1427990Min: 1984310 / Avg: 1990790 / Max: 19996801. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Neural Magic DeepSparse

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On150300450600750SE +/- 0.92, N = 3SE +/- 0.18, N = 3SE +/- 0.51, N = 3SE +/- 1.22, N = 3SE +/- 1.03, N = 3SE +/- 5.42, N = 13158.53221.73232.21355.67491.34704.58
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On120240360480600Min: 156.79 / Avg: 158.53 / Max: 159.93Min: 221.44 / Avg: 221.73 / Max: 222.07Min: 231.21 / Avg: 232.21 / Max: 232.86Min: 353.37 / Avg: 355.67 / Max: 357.52Min: 489.36 / Avg: 491.34 / Max: 492.82Min: 681.88 / Avg: 704.58 / Max: 762.39

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2AVX-512 OffAVX-512 On2K4K6K8K10KSE +/- 12.20, N = 3SE +/- 13.49, N = 3SE +/- 24.70, N = 6SE +/- 9.11, N = 3SE +/- 46.49, N = 3SE +/- 47.16, N = 41993.072588.762602.653643.846371.328562.691. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2AVX-512 OffAVX-512 On15003000450060007500Min: 1972.67 / Avg: 1993.07 / Max: 2014.86Min: 2573.18 / Avg: 2588.76 / Max: 2615.62Min: 2485.24 / Avg: 2602.65 / Max: 2648.84Min: 3628.3 / Avg: 3643.84 / Max: 3659.84Min: 6283.52 / Avg: 6371.32 / Max: 6441.74Min: 8421.46 / Avg: 8562.69 / Max: 8617.61. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2AVX-512 OffAVX-512 On70140210280350SE +/- 0.49, N = 3SE +/- 0.54, N = 3SE +/- 0.99, N = 6SE +/- 0.36, N = 3SE +/- 1.86, N = 3SE +/- 1.89, N = 479.72103.55104.11145.75254.85342.511. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgBillion Interactions/s Per Watt, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2AVX-512 OffAVX-512 On0.15280.30560.45840.61120.7640.1560.2040.1700.2460.4810.679
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterminiBUDE 20210901CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2943.56 / Max: 3400Min: 2300 / Avg: 2927.95 / Max: 3402Min: 1900 / Avg: 3016.33 / Max: 3515Min: 1900 / Avg: 2986.92 / Max: 3558Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterminiBUDE 20210901CPU Power Consumption MonitorAVX-512 OffAVX-512 On150300450600750Min: 114.01 / Avg: 512.58 / Max: 652.62Min: 122.79 / Avg: 508.57 / Max: 665.4Min: 109.04 / Avg: 612 / Max: 868.4Min: 209.21 / Avg: 593.08 / Max: 864.78Min: 54.45 / Avg: 530.15 / Max: 665.49Min: 59.34 / Avg: 504.56 / Max: 653.44
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterminiBUDE 20210901CPU Temperature MonitorAVX-512 OffAVX-512 On1632486480Min: 46 / Avg: 72.36 / Max: 85Min: 47 / Avg: 72.68 / Max: 84Min: 37 / Avg: 52.39 / Max: 59Min: 37 / Avg: 48.68 / Max: 55Min: 34.13 / Avg: 48.71 / Max: 55.63Min: 34.13 / Avg: 47.83 / Max: 54.25
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2AVX-512 OffAVX-512 On60120180240300Min: 78.91 / Avg: 79.72 / Max: 80.59Min: 102.93 / Avg: 103.55 / Max: 104.63Min: 99.41 / Avg: 104.11 / Max: 105.95Min: 145.13 / Avg: 145.75 / Max: 146.39Min: 251.34 / Avg: 254.85 / Max: 257.67Min: 336.86 / Avg: 342.51 / Max: 344.71. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: scryptAVX-512 OffAVX-512 On10002000300040005000SE +/- 0.63, N = 3SE +/- 7.11, N = 3SE +/- 0.54, N = 3SE +/- 1.21, N = 3SE +/- 0.03, N = 3SE +/- 13.77, N = 31122.742307.421623.883665.552947.944816.311. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s Per Watt, More Is BetterCpuminer-Opt 3.20.3Algorithm: scryptAVX-512 OffAVX-512 On2468102.2494.6272.6185.6365.4378.054
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterCpuminer-Opt 3.20.3CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2914.43 / Max: 3000Min: 2300 / Avg: 2665.18 / Max: 2970Min: 1900 / Avg: 2927.59 / Max: 3501Min: 1900 / Avg: 2817.46 / Max: 3508Min: 2936 / Avg: 3110.3 / Max: 3709Min: 2986 / Avg: 3249.64 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterCpuminer-Opt 3.20.3CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 123.97 / Avg: 499.12 / Max: 602.56Min: 114.85 / Avg: 498.71 / Max: 604.01Min: 108.05 / Avg: 620.23 / Max: 772.37Min: 205.47 / Avg: 650.33 / Max: 782.01Min: 52.42 / Avg: 542.18 / Max: 637.99Min: 59.62 / Avg: 597.99 / Max: 707.37
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterCpuminer-Opt 3.20.3CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 49 / Avg: 68.06 / Max: 73Min: 48 / Avg: 68.72 / Max: 75Min: 37 / Avg: 50.54 / Max: 55Min: 36 / Avg: 51.1 / Max: 56Min: 34.13 / Avg: 48.03 / Max: 52Min: 34.13 / Avg: 50.14 / Max: 55.63
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: scryptAVX-512 OffAVX-512 On8001600240032004000Min: 1121.73 / Avg: 1122.74 / Max: 1123.89Min: 2299.47 / Avg: 2307.42 / Max: 2321.6Min: 1622.99 / Avg: 1623.88 / Max: 1624.84Min: 3663.29 / Avg: 3665.55 / Max: 3667.44Min: 2947.91 / Avg: 2947.94 / Max: 2947.99Min: 4796.1 / Avg: 4816.31 / Max: 4842.611. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: x25xAVX-512 OffAVX-512 On16003200480064008000SE +/- 4.47, N = 3SE +/- 4.49, N = 3SE +/- 21.26, N = 6SE +/- 26.18, N = 3SE +/- 0.55, N = 3SE +/- 54.56, N = 31840.412650.362166.904071.595712.587691.781. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s Per Watt, More Is BetterCpuminer-Opt 3.20.3Algorithm: x25xAVX-512 OffAVX-512 On36912153.7135.3383.3136.32210.30213.447
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterCpuminer-Opt 3.20.3CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2916.53 / Max: 3305Min: 2300 / Avg: 2832.8 / Max: 3402Min: 1900 / Avg: 2906.57 / Max: 3511Min: 1900 / Avg: 2729.58 / Max: 3501Min: 3295 / Avg: 3487.56 / Max: 3709Min: 3132 / Avg: 3368.18 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterCpuminer-Opt 3.20.3CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 118.75 / Avg: 495.69 / Max: 593.22Min: 128.03 / Avg: 496.51 / Max: 596.23Min: 137.07 / Avg: 654.01 / Max: 765.2Min: 210.54 / Avg: 644.01 / Max: 768.58Min: 60.23 / Avg: 554.51 / Max: 659.93Min: 55.42 / Avg: 572.01 / Max: 668.33
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterCpuminer-Opt 3.20.3CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 53 / Avg: 69.68 / Max: 74Min: 51 / Avg: 69.47 / Max: 74Min: 40 / Avg: 53.43 / Max: 57Min: 38 / Avg: 52.08 / Max: 56Min: 36.75 / Avg: 50.8 / Max: 54.63Min: 37.25 / Avg: 52.15 / Max: 55.5
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: x25xAVX-512 OffAVX-512 On13002600390052006500Min: 1831.82 / Avg: 1840.41 / Max: 1846.87Min: 2641.98 / Avg: 2650.36 / Max: 2657.36Min: 2118.64 / Avg: 2166.9 / Max: 2265.53Min: 4032.79 / Avg: 4071.59 / Max: 4121.45Min: 5711.72 / Avg: 5712.58 / Max: 5713.61Min: 7590.41 / Avg: 7691.78 / Max: 7777.441. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1AVX-512 OffAVX-512 On60120180240300SE +/- 0.14, N = 5SE +/- 0.25, N = 5SE +/- 0.27, N = 6SE +/- 0.53, N = 7SE +/- 1.12, N = 8SE +/- 0.38, N = 1071.8593.5799.23143.45207.70290.961. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgBillion Interactions/s Per Watt, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1AVX-512 OffAVX-512 On0.23040.46080.69120.92161.1520.1610.2180.1810.2920.6191.024
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterminiBUDE 20210901CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2689.33 / Max: 3014Min: 2300 / Avg: 2694.51 / Max: 3194Min: 1900 / Avg: 2980.48 / Max: 3511Min: 1900 / Avg: 3105.81 / Max: 3512Min: 3145 / Avg: 3570.92 / Max: 3709Min: 3296 / Avg: 3597.25 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterminiBUDE 20210901CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 112.89 / Avg: 447.53 / Max: 605.44Min: 76.52 / Avg: 430.19 / Max: 621.53Min: 204.3 / Avg: 549.62 / Max: 791.37Min: 111.28 / Avg: 491.9 / Max: 790.45Min: 50.93 / Avg: 335.56 / Max: 660.26Min: 60.57 / Avg: 284.05 / Max: 711.18
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterminiBUDE 20210901CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 50 / Avg: 65.07 / Max: 72Min: 52 / Avg: 65.79 / Max: 74Min: 39 / Avg: 48.48 / Max: 54Min: 40 / Avg: 46.05 / Max: 51Min: 35 / Avg: 42.25 / Max: 50.38Min: 35.88 / Avg: 41.15 / Max: 49
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1AVX-512 OffAVX-512 On50100150200250Min: 71.33 / Avg: 71.85 / Max: 72.1Min: 92.91 / Avg: 93.57 / Max: 94.14Min: 98.52 / Avg: 99.22 / Max: 100.15Min: 141.83 / Avg: 143.45 / Max: 145.77Min: 204.62 / Avg: 207.7 / Max: 214.13Min: 288.45 / Avg: 290.96 / Max: 292.631. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1AVX-512 OffAVX-512 On16003200480064008000SE +/- 3.51, N = 5SE +/- 6.26, N = 5SE +/- 6.63, N = 6SE +/- 13.13, N = 7SE +/- 27.90, N = 8SE +/- 9.60, N = 101796.162339.322480.613586.275192.437274.051. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1AVX-512 OffAVX-512 On13002600390052006500Min: 1783.27 / Avg: 1796.16 / Max: 1802.48Min: 2322.63 / Avg: 2339.32 / Max: 2353.47Min: 2463 / Avg: 2480.61 / Max: 2503.65Min: 3545.69 / Avg: 3586.27 / Max: 3644.17Min: 5115.49 / Avg: 5192.43 / Max: 5353.21Min: 7211.16 / Avg: 7274.05 / Max: 7315.831. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

Neural Magic DeepSparse

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On400800120016002000SE +/- 1.47, N = 3SE +/- 0.98, N = 3SE +/- 0.84, N = 3SE +/- 1.88, N = 3SE +/- 0.94, N = 3SE +/- 1.71, N = 3541.86835.78832.851492.881410.421966.41
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On30060090012001500Min: 539.25 / Avg: 541.86 / Max: 544.36Min: 833.96 / Avg: 835.78 / Max: 837.31Min: 831.73 / Avg: 832.85 / Max: 834.49Min: 1490.44 / Avg: 1492.88 / Max: 1496.57Min: 1409.02 / Avg: 1410.42 / Max: 1412.22Min: 1963.17 / Avg: 1966.41 / Max: 1969

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On130260390520650SE +/- 0.52, N = 3SE +/- 0.49, N = 3SE +/- 0.93, N = 3SE +/- 0.95, N = 3SE +/- 0.22, N = 3SE +/- 0.55, N = 3183.83227.45264.77424.27510.88613.94
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On110220330440550Min: 183.04 / Avg: 183.83 / Max: 184.81Min: 226.64 / Avg: 227.45 / Max: 228.32Min: 263.04 / Avg: 264.77 / Max: 266.21Min: 422.52 / Avg: 424.27 / Max: 425.77Min: 510.43 / Avg: 510.88 / Max: 511.15Min: 612.98 / Avg: 613.94 / Max: 614.87

Mobile Neural Network

MNN is the Mobile Neural Network as a highly efficient, lightweight deep learning framework developed by Alibaba. This MNN test profile is building the OpenMP / CPU threaded version for processor benchmarking and not any GPU-accelerated test. MNN does allow making use of AVX-512 extensions. Learn more via the OpenBenchmarking.org test page.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.1Model: squeezenetv1.1AVX-512 OffAVX-512 On246810SE +/- 0.039, N = 10SE +/- 0.064, N = 15SE +/- 0.008, N = 3SE +/- 0.072, N = 3SE +/- 0.137, N = 3SE +/- 0.219, N = 92.6782.2973.5512.7377.5667.2041. (CXX) g++ options: -O3 -march=native -mno-amx-tile -std=c++11 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.1Model: squeezenetv1.1AVX-512 OffAVX-512 On3691215Min: 2.55 / Avg: 2.68 / Max: 2.87Min: 2.02 / Avg: 2.3 / Max: 2.61Min: 3.54 / Avg: 3.55 / Max: 3.56Min: 2.6 / Avg: 2.74 / Max: 2.84Min: 7.37 / Avg: 7.57 / Max: 7.83Min: 6.21 / Avg: 7.2 / Max: 8.021. (CXX) g++ options: -O3 -march=native -mno-amx-tile -std=c++11 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Neural Magic DeepSparse

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Detection,YOLOv5s COCO - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On2004006008001000SE +/- 0.62, N = 3SE +/- 0.38, N = 3SE +/- 0.43, N = 3SE +/- 0.33, N = 3SE +/- 0.26, N = 3SE +/- 2.56, N = 3247.27318.79405.72592.45660.54809.08
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Detection,YOLOv5s COCO - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On140280420560700Min: 246.52 / Avg: 247.27 / Max: 248.51Min: 318.03 / Avg: 318.79 / Max: 319.19Min: 405.1 / Avg: 405.72 / Max: 406.53Min: 591.98 / Avg: 592.45 / Max: 593.1Min: 660.03 / Avg: 660.54 / Max: 660.85Min: 803.95 / Avg: 809.08 / Max: 811.76

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Medium Mesh Size - Execution TimeAVX-512 OffAVX-512 On80160240320400358.63361.37194.79195.48110.95112.381. (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
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOpenFOAM 10CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2959.08 / Max: 3400Min: 2300 / Avg: 2948.55 / Max: 3400Min: 1900 / Avg: 2735.03 / Max: 3517Min: 1900 / Avg: 2718.73 / Max: 3515Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709.02 / Max: 3714
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOpenFOAM 10CPU Power Consumption MonitorAVX-512 OffAVX-512 On130260390520650Min: 122.02 / Avg: 539.74 / Max: 590.97Min: 102.2 / Avg: 540.67 / Max: 588.28Min: 211.03 / Avg: 696.05 / Max: 761.31Min: 210.74 / Avg: 695.05 / Max: 754.74Min: 57.64 / Avg: 661.82 / Max: 737.11Min: 61.7 / Avg: 659.15 / Max: 738.72
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOpenFOAM 10CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 39 / Avg: 71.62 / Max: 78Min: 40 / Avg: 71.85 / Max: 80Min: 30 / Avg: 53.51 / Max: 59Min: 33 / Avg: 54.15 / Max: 59Min: 31.88 / Avg: 53.96 / Max: 60Min: 36.38 / Avg: 55.35 / Max: 60.88

Neural Magic DeepSparse

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On30060090012001500SE +/- 0.62, N = 3SE +/- 0.94, N = 3SE +/- 0.51, N = 3SE +/- 2.15, N = 3SE +/- 0.56, N = 3SE +/- 3.78, N = 3379.58455.03584.15875.641006.081205.18
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On2004006008001000Min: 378.95 / Avg: 379.58 / Max: 380.81Min: 453.31 / Avg: 455.03 / Max: 456.55Min: 583.21 / Avg: 584.15 / Max: 584.96Min: 871.69 / Avg: 875.64 / Max: 879.1Min: 1004.96 / Avg: 1006.08 / Max: 1006.68Min: 1200.58 / Avg: 1205.18 / Max: 1212.68

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.10Benchmark: gravity_spheres_volume/dim_512/scivis/real_timeAVX-512 OffAVX-512 On1020304050SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.08, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.10, N = 313.7821.0622.8335.9525.3143.24
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgItems Per Second Per Watt, More Is BetterOSPRay 2.10Benchmark: gravity_spheres_volume/dim_512/scivis/real_timeAVX-512 OffAVX-512 On0.01670.03340.05010.06680.08350.0260.0400.0330.0530.0410.074
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOSPRay 2.10CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2842.78 / Max: 3400Min: 2200 / Avg: 2873.43 / Max: 3400Min: 1900 / Avg: 2870.01 / Max: 3506Min: 1900 / Avg: 2689.91 / Max: 3502Min: 2771 / Avg: 3286.61 / Max: 3709Min: 2796 / Avg: 3267.78 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOSPRay 2.10CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 133.04 / Avg: 523.63 / Max: 645.71Min: 124.97 / Avg: 528.92 / Max: 632.61Min: 206.88 / Avg: 688.18 / Max: 807.45Min: 211.39 / Avg: 674.63 / Max: 797.32Min: 57.79 / Avg: 616.9 / Max: 718.46Min: 60.18 / Avg: 581.84 / Max: 672.96
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOSPRay 2.10CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 56 / Avg: 72.16 / Max: 76Min: 52 / Avg: 71.64 / Max: 75Min: 39 / Avg: 55.05 / Max: 58Min: 40 / Avg: 53.77 / Max: 57Min: 37.75 / Avg: 52.77 / Max: 55.5Min: 35.88 / Avg: 49.7 / Max: 52.88
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.10Benchmark: gravity_spheres_volume/dim_512/scivis/real_timeAVX-512 OffAVX-512 On918273645Min: 13.71 / Avg: 13.78 / Max: 13.82Min: 21.02 / Avg: 21.06 / Max: 21.11Min: 22.69 / Avg: 22.83 / Max: 22.98Min: 35.9 / Avg: 35.95 / Max: 36Min: 25.26 / Avg: 25.31 / Max: 25.36Min: 43.05 / Avg: 43.24 / Max: 43.35

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.10Benchmark: gravity_spheres_volume/dim_512/ao/real_timeAVX-512 OffAVX-512 On1020304050SE +/- 0.05, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.12, N = 3SE +/- 0.04, N = 3SE +/- 0.06, N = 314.3521.2223.1736.3626.4043.71
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgItems Per Second Per Watt, More Is BetterOSPRay 2.10Benchmark: gravity_spheres_volume/dim_512/ao/real_timeAVX-512 OffAVX-512 On0.01690.03380.05070.06760.08450.0270.0400.0340.0540.0430.075
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOSPRay 2.10CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2855.45 / Max: 3400Min: 2300 / Avg: 2866.53 / Max: 3402Min: 1900 / Avg: 2900.68 / Max: 3504Min: 1900 / Avg: 2652.54 / Max: 3514Min: 2771 / Avg: 3210.13 / Max: 3709Min: 2802 / Avg: 3315.52 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOSPRay 2.10CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 119.36 / Avg: 524.83 / Max: 617.87Min: 119.41 / Avg: 530.36 / Max: 641.9Min: 107.51 / Avg: 689.16 / Max: 818.74Min: 209.17 / Avg: 676.18 / Max: 801.55Min: 56.53 / Avg: 614.11 / Max: 718.03Min: 56.1 / Avg: 580.71 / Max: 668.08
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOSPRay 2.10CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 46 / Avg: 72.76 / Max: 79Min: 46 / Avg: 73.63 / Max: 80Min: 35 / Avg: 54.32 / Max: 58Min: 34 / Avg: 52.96 / Max: 57Min: 34.63 / Avg: 51.95 / Max: 55.25Min: 32.75 / Avg: 49.32 / Max: 52
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.10Benchmark: gravity_spheres_volume/dim_512/ao/real_timeAVX-512 OffAVX-512 On918273645Min: 14.25 / Avg: 14.35 / Max: 14.4Min: 21.15 / Avg: 21.22 / Max: 21.27Min: 23.1 / Avg: 23.17 / Max: 23.24Min: 36.17 / Avg: 36.36 / Max: 36.57Min: 26.35 / Avg: 26.4 / Max: 26.48Min: 43.61 / Avg: 43.71 / Max: 43.81

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16 - Device: CPUAVX-512 OffAVX-512 On510152025SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 319.6817.8317.298.0112.646.461. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fsigned-char -ffunction-sections -fdata-sections -fno-strict-overflow -fwrapv
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2022.3CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2788.99 / Max: 3402Min: 2300 / Avg: 2555.76 / Max: 3400Min: 1900 / Avg: 2479.96 / Max: 3506Min: 1900 / Avg: 2431.39 / Max: 3509Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 121.86 / Avg: 518.5 / Max: 595.33Min: 124.75 / Avg: 518.86 / Max: 596.78Min: 208.74 / Avg: 674.93 / Max: 776.16Min: 209.62 / Avg: 674 / Max: 775.31Min: 54.82 / Avg: 604.27 / Max: 663.05Min: 38.84 / Avg: 638.51 / Max: 699.19
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOpenVINO 2022.3CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 54 / Avg: 72.7 / Max: 77Min: 54 / Avg: 72.62 / Max: 77Min: 40 / Avg: 53.53 / Max: 57Min: 40 / Avg: 52.8 / Max: 55Min: 37.25 / Avg: 50.88 / Max: 53.5Min: 37.13 / Avg: 52.63 / Max: 55.5
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Vehicle Detection FP16 - Device: CPUAVX-512 OffAVX-512 On510152025Min: 19.65 / Avg: 19.68 / Max: 19.69Min: 17.8 / Avg: 17.83 / Max: 17.85Min: 17.28 / Avg: 17.29 / Max: 17.3Min: 8.01 / Avg: 8.01 / Max: 8.02Min: 12.62 / Avg: 12.64 / Max: 12.65Min: 6.46 / Avg: 6.46 / Max: 6.461. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fsigned-char -ffunction-sections -fdata-sections -fno-strict-overflow -fwrapv

Mobile Neural Network

MNN is the Mobile Neural Network as a highly efficient, lightweight deep learning framework developed by Alibaba. This MNN test profile is building the OpenMP / CPU threaded version for processor benchmarking and not any GPU-accelerated test. MNN does allow making use of AVX-512 extensions. Learn more via the OpenBenchmarking.org test page.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.1Model: resnet-v2-50AVX-512 OffAVX-512 On612182430SE +/- 0.099, N = 10SE +/- 0.074, N = 15SE +/- 0.060, N = 3SE +/- 0.065, N = 3SE +/- 0.300, N = 3SE +/- 0.193, N = 918.1337.90917.7168.49624.08114.7271. (CXX) g++ options: -O3 -march=native -mno-amx-tile -std=c++11 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.1Model: resnet-v2-50AVX-512 OffAVX-512 On612182430Min: 17.43 / Avg: 18.13 / Max: 18.56Min: 7.54 / Avg: 7.91 / Max: 8.34Min: 17.6 / Avg: 17.72 / Max: 17.78Min: 8.37 / Avg: 8.5 / Max: 8.59Min: 23.52 / Avg: 24.08 / Max: 24.55Min: 13.56 / Avg: 14.73 / Max: 15.391. (CXX) g++ options: -O3 -march=native -mno-amx-tile -std=c++11 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs 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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer ISPC - Model: CrownAVX-512 OffAVX-512 On4080120160200SE +/- 0.27, N = 5SE +/- 0.17, N = 5SE +/- 5.25, N = 15SE +/- 0.53, N = 6SE +/- 0.42, N = 7SE +/- 0.47, N = 870.5174.8062.0492.65153.18181.89
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterEmbree 3.13Binary: Pathtracer ISPC - Model: CrownAVX-512 OffAVX-512 On0.11450.2290.34350.4580.57250.1670.1840.1100.1930.4180.509
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterEmbree 3.13CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2850.31 / Max: 3400Min: 2300 / Avg: 2636.37 / Max: 3400Min: 1900 / Avg: 3070.52 / Max: 3512Min: 1900 / Avg: 3006.61 / Max: 3508Min: 2771 / Avg: 3462.8 / Max: 3709Min: 2763 / Avg: 3454.37 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterEmbree 3.13CPU Power Consumption MonitorAVX-512 OffAVX-512 On130260390520650Min: 80.21 / Avg: 421.57 / Max: 583.55Min: 118.24 / Avg: 406.24 / Max: 561.45Min: 172.58 / Avg: 565.19 / Max: 755.21Min: 209.36 / Avg: 478.88 / Max: 665.78Min: 42.7 / Avg: 366.77 / Max: 692.37Min: 36.29 / Avg: 357.33 / Max: 701.32
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterEmbree 3.13CPU Temperature MonitorAVX-512 OffAVX-512 On1326395265Min: 48 / Avg: 61.37 / Max: 69Min: 46 / Avg: 57.94 / Max: 65Min: 37 / Avg: 46.83 / Max: 51Min: 34 / Avg: 42.39 / Max: 48Min: 35 / Avg: 42.18 / Max: 48.88Min: 33.25 / Avg: 39.98 / Max: 46.63
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer ISPC - Model: CrownAVX-512 OffAVX-512 On306090120150Min: 70.13 / Avg: 70.51 / Max: 71.57Min: 74.52 / Avg: 74.8 / Max: 75.49Min: 42.58 / Avg: 62.04 / Max: 90.41Min: 90.89 / Avg: 92.65 / Max: 94.17Min: 151.26 / Avg: 153.18 / Max: 154.52Min: 179.83 / Avg: 181.89 / Max: 183.6

Cpuminer-Opt

Cpuminer-Opt is a fork of cpuminer-multi that carries a wide range of CPU performance optimizations for measuring the potential cryptocurrency mining performance of the CPU/processor with a wide variety of cryptocurrencies. The benchmark reports the hash speed for the CPU mining performance for the selected cryptocurrency. Learn more via the OpenBenchmarking.org test page.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: DeepcoinAVX-512 OffAVX-512 On40K80K120K160K200KSE +/- 32.15, N = 3SE +/- 52.92, N = 3SE +/- 928.66, N = 3SE +/- 155.06, N = 3SE +/- 204.21, N = 3SE +/- 536.94, N = 35793064690911071055131652401602901. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s Per Watt, More Is BetterCpuminer-Opt 3.20.3Algorithm: DeepcoinAVX-512 OffAVX-512 On60120180240300116.45130.17139.12161.98284.10274.20
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterCpuminer-Opt 3.20.3CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2798.55 / Max: 3145Min: 2300 / Avg: 2779.27 / Max: 3400Min: 1900 / Avg: 2857.54 / Max: 3507Min: 1900 / Avg: 2936.48 / Max: 3508Min: 2899 / Avg: 3360.04 / Max: 3709Min: 3239 / Avg: 3515.22 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterCpuminer-Opt 3.20.3CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 88.2 / Avg: 497.46 / Max: 601.66Min: 71.69 / Avg: 496.96 / Max: 602.54Min: 212.39 / Avg: 654.88 / Max: 782.5Min: 202.83 / Avg: 651.38 / Max: 782.3Min: 60.67 / Avg: 581.62 / Max: 686.17Min: 31.65 / Avg: 584.57 / Max: 690.12
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterCpuminer-Opt 3.20.3CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 53 / Avg: 70.18 / Max: 74Min: 52 / Avg: 70.34 / Max: 75Min: 42 / Avg: 54.53 / Max: 58Min: 39 / Avg: 53.02 / Max: 57Min: 36.38 / Avg: 50.56 / Max: 53.75Min: 36.63 / Avg: 50.78 / Max: 54.25
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgkH/s, More Is BetterCpuminer-Opt 3.20.3Algorithm: DeepcoinAVX-512 OffAVX-512 On30K60K90K120K150KMin: 57880 / Avg: 57930 / Max: 57990Min: 64610 / Avg: 64690 / Max: 64790Min: 89820 / Avg: 91106.67 / Max: 92910Min: 105320 / Avg: 105513.33 / Max: 105820Min: 164940 / Avg: 165240 / Max: 165630Min: 159240 / Avg: 160290 / Max: 1610101. (CXX) g++ options: -O3 -march=native -mno-amx-tile -lcurl -lz -lpthread -lssl -lcrypto -lgmp

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 3 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path TracerAVX-512 OffAVX-512 On13K26K39K52K65KSE +/- 56.19, N = 3SE +/- 4.98, N = 3SE +/- 115.86, N = 3SE +/- 45.83, N = 3SE +/- 18.25, N = 3SE +/- 44.52, N = 36278049847373003274926541221201. (CXX) g++ options: -O3 -march=native -mno-amx-tile -ldl
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOSPRay Studio 0.11CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 1597 / Avg: 2940.27 / Max: 3407Min: 2300 / Avg: 2577.91 / Max: 3401Min: 1900 / Avg: 2875.06 / Max: 3514Min: 1900 / Avg: 2803.99 / Max: 3513Min: 2721 / Avg: 3353.6 / Max: 3709Min: 2702 / Avg: 3343.21 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorAVX-512 OffAVX-512 On150300450600750Min: 123.88 / Avg: 460.86 / Max: 660.32Min: 102.13 / Avg: 519.61 / Max: 662.67Min: 180.6 / Avg: 673.37 / Max: 881.95Min: 211.24 / Avg: 659.15 / Max: 861.5Min: 31.18 / Avg: 618.56 / Max: 788.32Min: 61.78 / Avg: 592.07 / Max: 769.69
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOSPRay Studio 0.11CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 53 / Avg: 66.34 / Max: 73Min: 49 / Avg: 70.06 / Max: 75Min: 39 / Avg: 53.34 / Max: 58Min: 39 / Avg: 52.54 / Max: 57Min: 38.13 / Avg: 52.62 / Max: 59.25Min: 36.75 / Avg: 51.43 / Max: 54.63
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 3 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path TracerAVX-512 OffAVX-512 On11K22K33K44K55KMin: 62670 / Avg: 62780.33 / Max: 62854Min: 49838 / Avg: 49847.33 / Max: 49855Min: 37088 / Avg: 37300 / Max: 37487Min: 32659 / Avg: 32749.33 / Max: 32808Min: 26508 / Avg: 26541 / Max: 26571Min: 22055 / Avg: 22119.67 / Max: 222051. (CXX) g++ options: -O3 -march=native -mno-amx-tile -ldl

Neural Magic DeepSparse

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On306090120150SE +/- 0.10, N = 3SE +/- 0.21, N = 3SE +/- 0.17, N = 3SE +/- 0.03, N = 3SE +/- 0.22, N = 3SE +/- 0.19, N = 343.1148.4166.7592.02120.5184.87
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On20406080100Min: 42.95 / Avg: 43.11 / Max: 43.28Min: 48.2 / Avg: 48.41 / Max: 48.83Min: 66.5 / Avg: 66.75 / Max: 67.08Min: 91.99 / Avg: 92.02 / Max: 92.08Min: 120.09 / Avg: 120.51 / Max: 120.77Min: 84.52 / Avg: 84.87 / Max: 85.16

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On306090120150SE +/- 0.16, N = 3SE +/- 0.01, N = 3SE +/- 0.23, N = 3SE +/- 0.19, N = 3SE +/- 0.10, N = 3SE +/- 0.10, N = 343.1648.2466.7392.03120.5284.40
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On20406080100Min: 42.89 / Avg: 43.16 / Max: 43.43Min: 48.23 / Avg: 48.24 / Max: 48.26Min: 66.29 / Avg: 66.73 / Max: 67.09Min: 91.66 / Avg: 92.03 / Max: 92.23Min: 120.37 / Avg: 120.52 / Max: 120.7Min: 84.2 / Avg: 84.4 / Max: 84.51

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.0Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPUAVX-512 OffAVX-512 On0.45280.90561.35841.81122.264SE +/- 0.002804, N = 9SE +/- 0.001297, N = 9SE +/- 0.006513, N = 9SE +/- 0.001579, N = 9SE +/- 0.002552, N = 9SE +/- 0.003984, N = 91.9928600.8751962.0124100.7368221.7366700.9862141. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetteroneDNN 3.0CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2610.75 / Max: 3145Min: 2300 / Avg: 2463.92 / Max: 3025Min: 1900 / Avg: 3132.72 / Max: 3511Min: 1900 / Avg: 3013.2 / Max: 3507Min: 3570 / Avg: 3666 / Max: 3709Min: 3403 / Avg: 3617.61 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetteroneDNN 3.0CPU Power Consumption MonitorAVX-512 OffAVX-512 On150300450600750Min: 117.63 / Avg: 306.45 / Max: 584.73Min: 120.33 / Avg: 330.6 / Max: 647.31Min: 210.1 / Avg: 417.66 / Max: 769.29Min: 204.47 / Avg: 439.83 / Max: 838.42Min: 56.11 / Avg: 241.2 / Max: 540.52Min: 30.68 / Avg: 261.28 / Max: 580.04
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetteroneDNN 3.0CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 50 / Avg: 57.62 / Max: 69Min: 51 / Avg: 59.32 / Max: 70Min: 33 / Avg: 39.82 / Max: 48Min: 36 / Avg: 40.95 / Max: 47Min: 33.63 / Avg: 36.99 / Max: 40.88Min: 34.13 / Avg: 37.73 / Max: 41.75
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.0Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPUAVX-512 OffAVX-512 On246810Min: 1.98 / Avg: 1.99 / Max: 2.01Min: 0.87 / Avg: 0.88 / Max: 0.88Min: 1.99 / Avg: 2.01 / Max: 2.05Min: 0.73 / Avg: 0.74 / Max: 0.74Min: 1.73 / Avg: 1.74 / Max: 1.75Min: 0.97 / Avg: 0.99 / Max: 11. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.10Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_timeAVX-512 OffAVX-512 On1224364860SE +/- 0.03, N = 3SE +/- 0.11, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 319.9625.5232.4842.9740.6754.16
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgItems Per Second Per Watt, More Is BetterOSPRay 2.10Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_timeAVX-512 OffAVX-512 On0.02070.04140.06210.08280.10350.0380.0490.0480.0640.0650.092
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOSPRay 2.10CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On7001400210028003500Min: 2293 / Avg: 2874.31 / Max: 3400Min: 2300 / Avg: 2590.11 / Max: 3402Min: 1900 / Avg: 2827.79 / Max: 3508Min: 1900 / Avg: 2769.32 / Max: 3512Min: 2690 / Avg: 3283.87 / Max: 3709Min: 2764 / Avg: 3287.4 / Max: 3772
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOSPRay 2.10CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 122.95 / Avg: 530.21 / Max: 639.78Min: 80.49 / Avg: 523.67 / Max: 629.96Min: 205.85 / Avg: 679.41 / Max: 796.02Min: 210.52 / Avg: 667.18 / Max: 772.19Min: 61.53 / Avg: 624.39 / Max: 724.31Min: 60.05 / Avg: 591.62 / Max: 680.48
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOSPRay 2.10CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 55 / Avg: 73.2 / Max: 77Min: 53 / Avg: 71.52 / Max: 75Min: 40 / Avg: 55.09 / Max: 58Min: 39 / Avg: 54.1 / Max: 58Min: 37.25 / Avg: 53.21 / Max: 56.88Min: 36.38 / Avg: 50.52 / Max: 53.25
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 2.10Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_timeAVX-512 OffAVX-512 On1122334455Min: 19.91 / Avg: 19.96 / Max: 20.01Min: 25.38 / Avg: 25.52 / Max: 25.74Min: 32.46 / Avg: 32.48 / Max: 32.54Min: 42.91 / Avg: 42.97 / Max: 43.06Min: 40.58 / Avg: 40.67 / Max: 40.75Min: 54.12 / Avg: 54.16 / Max: 54.19

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 1 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path TracerAVX-512 OffAVX-512 On10K20K30K40K50KSE +/- 10.69, N = 3SE +/- 44.00, N = 3SE +/- 69.83, N = 3SE +/- 11.72, N = 3SE +/- 50.78, N = 3SE +/- 30.61, N = 34816541834313732720422475185001. (CXX) g++ options: -O3 -march=native -mno-amx-tile -ldl
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOSPRay Studio 0.11CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 1740 / Avg: 2881.42 / Max: 3403Min: 2300 / Avg: 2594.91 / Max: 3401Min: 1900 / Avg: 2877.33 / Max: 3513Min: 1900 / Avg: 2787.1 / Max: 3505Min: 2789 / Avg: 3177.76 / Max: 3709Min: 2771 / Avg: 3169.22 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorAVX-512 OffAVX-512 On160320480640800Min: 130.23 / Avg: 517.4 / Max: 652.59Min: 101.37 / Avg: 511.05 / Max: 668.12Min: 209.5 / Avg: 662.67 / Max: 884.59Min: 114.26 / Avg: 669.91 / Max: 876.36Min: 58.52 / Avg: 614.53 / Max: 802.45Min: 32.83 / Avg: 631.84 / Max: 775.3
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOSPRay Studio 0.11CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 53 / Avg: 70.9 / Max: 77Min: 50 / Avg: 69.76 / Max: 75Min: 39 / Avg: 52.35 / Max: 57Min: 39 / Avg: 52.99 / Max: 57Min: 37.75 / Avg: 52.06 / Max: 55.5Min: 36.75 / Avg: 52.26 / Max: 54.63
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 1 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path TracerAVX-512 OffAVX-512 On8K16K24K32K40KMin: 48146 / Avg: 48165 / Max: 48183Min: 41789 / Avg: 41834 / Max: 41922Min: 31270 / Avg: 31372.67 / Max: 31506Min: 27181 / Avg: 27204.33 / Max: 27218Min: 22378 / Avg: 22474.67 / Max: 22550Min: 18465 / Avg: 18500 / Max: 185611. (CXX) g++ options: -O3 -march=native -mno-amx-tile -ldl

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 1 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path TracerAVX-512 OffAVX-512 On5K10K15K20K25KSE +/- 42.81, N = 3SE +/- 8.41, N = 3SE +/- 37.53, N = 3SE +/- 19.41, N = 3SE +/- 8.39, N = 3SE +/- 7.00, N = 3239432087215601135891121692601. (CXX) g++ options: -O3 -march=native -mno-amx-tile -ldl
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOSPRay Studio 0.11CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 1631 / Avg: 2895.11 / Max: 3404Min: 2300 / Avg: 2639.14 / Max: 3403Min: 1900 / Avg: 2891.51 / Max: 3509Min: 1900 / Avg: 2766.18 / Max: 3509Min: 2773 / Avg: 3122.61 / Max: 3709Min: 2682 / Avg: 3157.82 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorAVX-512 OffAVX-512 On150300450600750Min: 129.76 / Avg: 497.33 / Max: 659.2Min: 122.89 / Avg: 488.82 / Max: 662.51Min: 206.26 / Avg: 650.1 / Max: 877.56Min: 214.67 / Avg: 684.96 / Max: 875.35Min: 61.33 / Avg: 648.16 / Max: 799.28Min: 56.6 / Avg: 627.16 / Max: 783.63
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOSPRay Studio 0.11CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 53 / Avg: 68.82 / Max: 75Min: 50 / Avg: 67.06 / Max: 74Min: 38 / Avg: 51.75 / Max: 57Min: 38 / Avg: 53.78 / Max: 58Min: 37.25 / Avg: 53.26 / Max: 58.25Min: 36.75 / Avg: 51.41 / Max: 55.25
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 1 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path TracerAVX-512 OffAVX-512 On4K8K12K16K20KMin: 23890 / Avg: 23943.33 / Max: 24028Min: 20858 / Avg: 20871.67 / Max: 20887Min: 15561 / Avg: 15601 / Max: 15676Min: 13567 / Avg: 13589.33 / Max: 13628Min: 11202 / Avg: 11216 / Max: 11231Min: 9249 / Avg: 9260 / Max: 92731. (CXX) g++ options: -O3 -march=native -mno-amx-tile -ldl

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 3 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path TracerAVX-512 OffAVX-512 On6K12K18K24K30KSE +/- 46.52, N = 3SE +/- 18.50, N = 3SE +/- 45.36, N = 3SE +/- 25.54, N = 3SE +/- 32.66, N = 3SE +/- 3.79, N = 32849024927186081633213341110581. (CXX) g++ options: -O3 -march=native -mno-amx-tile -ldl
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOSPRay Studio 0.11CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 1835 / Avg: 2867.88 / Max: 3402Min: 2300 / Avg: 2615.33 / Max: 3403Min: 1900 / Avg: 2836.9 / Max: 3512Min: 1900 / Avg: 2841.62 / Max: 3508Min: 2726 / Avg: 3186.89 / Max: 3709Min: 2777 / Avg: 3169.51 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorAVX-512 OffAVX-512 On160320480640800Min: 117.01 / Avg: 524.66 / Max: 656.47Min: 100.35 / Avg: 499.04 / Max: 664.34Min: 179.46 / Avg: 687.97 / Max: 887.42Min: 107.88 / Avg: 652.2 / Max: 894.12Min: 60.29 / Avg: 636.59 / Max: 798.92Min: 37.81 / Avg: 629.17 / Max: 778.51
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOSPRay Studio 0.11CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 51 / Avg: 71.52 / Max: 77Min: 52 / Avg: 68.49 / Max: 74Min: 38 / Avg: 53.76 / Max: 58Min: 39 / Avg: 51.56 / Max: 57Min: 37.25 / Avg: 52.98 / Max: 56.38Min: 37.25 / Avg: 52.28 / Max: 56.38
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 3 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path TracerAVX-512 OffAVX-512 On5K10K15K20K25KMin: 28401 / Avg: 28490 / Max: 28558Min: 24896 / Avg: 24927 / Max: 24960Min: 18528 / Avg: 18608.33 / Max: 18685Min: 16299 / Avg: 16331.67 / Max: 16382Min: 13301 / Avg: 13341.33 / Max: 13406Min: 11052 / Avg: 11058 / Max: 110651. (CXX) g++ options: -O3 -march=native -mno-amx-tile -ldl

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path TracerAVX-512 OffAVX-512 On30060090012001500SE +/- 1.00, N = 3SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 1.20, N = 3149113089808597055811. (CXX) g++ options: -O3 -march=native -mno-amx-tile -ldl
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOSPRay Studio 0.11CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 1783 / Avg: 2893.57 / Max: 3403Min: 2300 / Avg: 2622.27 / Max: 3401Min: 1900 / Avg: 2891.42 / Max: 3545Min: 1900 / Avg: 2818.86 / Max: 3514Min: 2720 / Avg: 3276 / Max: 3709Min: 2646 / Avg: 3229.97 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorAVX-512 OffAVX-512 On150300450600750Min: 129.57 / Avg: 500.65 / Max: 657.6Min: 102.33 / Avg: 497.27 / Max: 661.94Min: 212.33 / Avg: 663.86 / Max: 880.6Min: 178.24 / Avg: 661.77 / Max: 876.06Min: 60.36 / Avg: 615.21 / Max: 792.28Min: 60.79 / Avg: 611.59 / Max: 770.42
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOSPRay Studio 0.11CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 54 / Avg: 69.43 / Max: 76Min: 53 / Avg: 69.12 / Max: 75Min: 35 / Avg: 51.94 / Max: 57Min: 35 / Avg: 51.67 / Max: 57Min: 37.25 / Avg: 51.89 / Max: 55.5Min: 36.38 / Avg: 50.48 / Max: 53.75
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path TracerAVX-512 OffAVX-512 On30060090012001500Min: 1490 / Avg: 1491 / Max: 1493Min: 1308 / Avg: 1308.33 / Max: 1309Min: 979 / Avg: 979.67 / Max: 981Min: 858 / Avg: 858.67 / Max: 859Min: 704 / Avg: 704.67 / Max: 705Min: 579 / Avg: 580.67 / Max: 5831. (CXX) g++ options: -O3 -march=native -mno-amx-tile -ldl

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path TracerAVX-512 OffAVX-512 On400800120016002000SE +/- 1.00, N = 3SE +/- 0.00, N = 3SE +/- 1.20, N = 3SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 0.33, N = 317701563117010298366951. (CXX) g++ options: -O3 -march=native -mno-amx-tile -ldl
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOSPRay Studio 0.11CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 1947 / Avg: 2897.09 / Max: 3403Min: 2300 / Avg: 2609.05 / Max: 3401Min: 1900 / Avg: 2874.56 / Max: 3509Min: 1900 / Avg: 2823.98 / Max: 3514Min: 2721 / Avg: 3208.32 / Max: 3709Min: 2671 / Avg: 3308.27 / Max: 3711
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOSPRay Studio 0.11CPU Power Consumption MonitorAVX-512 OffAVX-512 On150300450600750Min: 81.76 / Avg: 503.22 / Max: 657.11Min: 111.41 / Avg: 500.78 / Max: 666.67Min: 212.51 / Avg: 666.51 / Max: 879.7Min: 214.02 / Avg: 660.08 / Max: 874.86Min: 60.27 / Avg: 620.19 / Max: 789.77Min: 58.16 / Avg: 599.55 / Max: 773.94
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOSPRay Studio 0.11CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 51 / Avg: 69.42 / Max: 76Min: 50 / Avg: 68.75 / Max: 75Min: 38 / Avg: 52.47 / Max: 57Min: 39 / Avg: 52.34 / Max: 57Min: 37.75 / Avg: 52.04 / Max: 55.5Min: 37.25 / Avg: 50.44 / Max: 53.75
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 0.11Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path TracerAVX-512 OffAVX-512 On30060090012001500Min: 1768 / Avg: 1770 / Max: 1771Min: 1563 / Avg: 1563 / Max: 1563Min: 1168 / Avg: 1170.33 / Max: 1172Min: 1028 / Avg: 1028.67 / Max: 1029Min: 835 / Avg: 836.33 / Max: 837Min: 694 / Avg: 694.67 / Max: 6951. (CXX) g++ options: -O3 -march=native -mno-amx-tile -ldl

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Large Mesh Size - Execution TimeAVX-512 OffAVX-512 On150030004500600075007106.767117.014319.124305.122803.542849.611. (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
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOpenFOAM 10CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2980.24 / Max: 3408Min: 2300 / Avg: 2974.56 / Max: 3400Min: 2491 / Avg: 2870.69 / Max: 3516Min: 2451 / Avg: 2868.19 / Max: 3513Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOpenFOAM 10CPU Power Consumption MonitorAVX-512 OffAVX-512 On130260390520650Min: 127.05 / Avg: 547.75 / Max: 588.4Min: 129.82 / Avg: 547.41 / Max: 588.74Min: 208.95 / Avg: 709.42 / Max: 767.54Min: 214.57 / Avg: 709.53 / Max: 762.94Min: 58.28 / Avg: 650.61 / Max: 748.65Min: 62.76 / Avg: 655.07 / Max: 742.42
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOpenFOAM 10CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 51 / Avg: 70.16 / Max: 75Min: 50 / Avg: 69.77 / Max: 75Min: 39 / Avg: 54.29 / Max: 60Min: 40 / Avg: 54.65 / Max: 60Min: 36.75 / Avg: 54.3 / Max: 61.38Min: 37.75 / Avg: 55.54 / Max: 62.63

Neural Magic DeepSparse

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On246810SE +/- 0.0346, N = 3SE +/- 0.0029, N = 3SE +/- 0.0060, N = 3SE +/- 0.0091, N = 3SE +/- 0.0009, N = 3SE +/- 0.0099, N = 38.81584.37505.32423.55094.84445.2240
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2943.87 / Max: 3400Min: 2300 / Avg: 2506.26 / Max: 3400Min: 1900 / Avg: 3287.49 / Max: 3517Min: 1900 / Avg: 3360.23 / Max: 3516Min: 3698 / Avg: 3708.91 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On100200300400500Min: 123.39 / Avg: 336.16 / Max: 377.19Min: 118.87 / Avg: 330.83 / Max: 360.91Min: 106.99 / Avg: 477.41 / Max: 537.73Min: 209.97 / Avg: 483.39 / Max: 522.36Min: 56.69 / Avg: 346.41 / Max: 397.2Min: 58.85 / Avg: 324.27 / Max: 371.22
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 55 / Avg: 72.43 / Max: 78Min: 53 / Avg: 69.21 / Max: 74Min: 36 / Avg: 47.99 / Max: 52Min: 38 / Avg: 48.31 / Max: 52Min: 43.5 / Avg: 54.68 / Max: 57.75Min: 43.5 / Avg: 51.97 / Max: 57.25
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On3691215Min: 8.77 / Avg: 8.82 / Max: 8.88Min: 4.37 / Avg: 4.38 / Max: 4.38Min: 5.31 / Avg: 5.32 / Max: 5.33Min: 3.54 / Avg: 3.55 / Max: 3.57Min: 4.84 / Avg: 4.84 / Max: 4.85Min: 5.2 / Avg: 5.22 / Max: 5.23

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On60120180240300SE +/- 0.44, N = 3SE +/- 0.16, N = 3SE +/- 0.21, N = 3SE +/- 0.71, N = 3SE +/- 0.04, N = 3SE +/- 0.36, N = 3113.33228.12187.65281.24206.32191.33
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On50100150200250Min: 112.48 / Avg: 113.33 / Max: 113.96Min: 227.91 / Avg: 228.12 / Max: 228.42Min: 187.33 / Avg: 187.65 / Max: 188.05Min: 279.83 / Avg: 281.24 / Max: 282.12Min: 206.24 / Avg: 206.32 / Max: 206.38Min: 190.96 / Avg: 191.33 / Max: 192.06

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs 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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer ISPC - Model: Asian DragonAVX-512 OffAVX-512 On50100150200250SE +/- 0.42, N = 6SE +/- 0.56, N = 6SE +/- 0.50, N = 6SE +/- 0.64, N = 6SE +/- 0.11, N = 8SE +/- 0.25, N = 996.34106.1890.40122.53177.68213.95
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterEmbree 3.13Binary: Pathtracer ISPC - Model: Asian DragonAVX-512 OffAVX-512 On0.15530.31060.46590.62120.77650.2400.2720.1680.2550.5180.690
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterEmbree 3.13CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2765.28 / Max: 3401Min: 2300 / Avg: 2565.14 / Max: 3400Min: 1900 / Avg: 3074.54 / Max: 3507Min: 1900 / Avg: 3148.38 / Max: 3511Min: 2833 / Avg: 3434 / Max: 3709Min: 2918 / Avg: 3474.26 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterEmbree 3.13CPU Power Consumption MonitorAVX-512 OffAVX-512 On130260390520650Min: 101.95 / Avg: 400.96 / Max: 615.31Min: 122.75 / Avg: 389.97 / Max: 625.41Min: 205.64 / Avg: 537.44 / Max: 760.22Min: 207.22 / Avg: 480.24 / Max: 708.89Min: 26.98 / Avg: 343.25 / Max: 711.13Min: 37.38 / Avg: 310.11 / Max: 703.12
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterEmbree 3.13CPU Temperature MonitorAVX-512 OffAVX-512 On1326395265Min: 44 / Avg: 57.26 / Max: 66Min: 50 / Avg: 59.72 / Max: 67Min: 39 / Avg: 46.46 / Max: 51Min: 34 / Avg: 41.38 / Max: 47Min: 35.88 / Avg: 41.18 / Max: 48Min: 32.38 / Avg: 37.98 / Max: 44.38
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer ISPC - Model: Asian DragonAVX-512 OffAVX-512 On4080120160200Min: 94.63 / Avg: 96.34 / Max: 97.51Min: 103.78 / Avg: 106.18 / Max: 107.77Min: 88.51 / Avg: 90.4 / Max: 92.2Min: 121.25 / Avg: 122.53 / Max: 125.37Min: 177.07 / Avg: 177.68 / Max: 178.23Min: 212.25 / Avg: 213.95 / Max: 214.74

LeelaChessZero

LeelaChessZero (lc0 / lczero) is a chess engine automated vian neural networks. This test profile can be used for OpenCL, CUDA + cuDNN, and BLAS (CPU-based) benchmarking. Learn more via the OpenBenchmarking.org test page.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.28Backend: EigenAVX-512 OffAVX-512 On2K4K6K8K10KSE +/- 58.86, N = 3SE +/- 64.32, N = 4SE +/- 63.14, N = 3SE +/- 151.11, N = 3SE +/- 66.84, N = 3SE +/- 35.63, N = 347095750824311134816190851. (CXX) g++ options: -flto -O3 -march=native -mno-amx-tile -pthread
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgNodes Per Second Per Watt, More Is BetterLeelaChessZero 0.28Backend: EigenAVX-512 OffAVX-512 On61218243012.0614.5315.3820.6121.8924.74
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterLeelaChessZero 0.28CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 3281.52 / Max: 3416Min: 2300 / Avg: 3283.63 / Max: 3411Min: 1900 / Avg: 3415.13 / Max: 3542Min: 1900 / Avg: 3431.65 / Max: 3563Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3711
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterLeelaChessZero 0.28CPU Power Consumption MonitorAVX-512 OffAVX-512 On130260390520650Min: 124.39 / Avg: 390.39 / Max: 591.33Min: 123.27 / Avg: 395.84 / Max: 585.79Min: 207.21 / Avg: 535.98 / Max: 730.58Min: 211.45 / Avg: 540.33 / Max: 766.98Min: 58.03 / Avg: 372.74 / Max: 519.8Min: 59.5 / Avg: 367.29 / Max: 503.22
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterLeelaChessZero 0.28CPU Temperature MonitorAVX-512 OffAVX-512 On1632486480Min: 44 / Avg: 65.29 / Max: 84Min: 45 / Avg: 66.24 / Max: 82Min: 35 / Avg: 47.75 / Max: 56Min: 34 / Avg: 43.65 / Max: 56Min: 34.63 / Avg: 46.31 / Max: 59.13Min: 33.25 / Avg: 40.64 / Max: 53.75
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.28Backend: EigenAVX-512 OffAVX-512 On2K4K6K8K10KMin: 4602 / Avg: 4709 / Max: 4805Min: 5617 / Avg: 5749.5 / Max: 5916Min: 8156 / Avg: 8243.33 / Max: 8366Min: 10891 / Avg: 11133.67 / Max: 11411Min: 8061 / Avg: 8161.33 / Max: 8288Min: 9033 / Avg: 9084.67 / Max: 91531. (CXX) g++ options: -flto -O3 -march=native -mno-amx-tile -pthread

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUAVX-512 OffAVX-512 On30K60K90K120K150KSE +/- 38.18, N = 3SE +/- 42.04, N = 3SE +/- 74.60, N = 3SE +/- 153.42, N = 3SE +/- 673.74, N = 3SE +/- 1008.13, N = 350413.9867701.6073906.76103184.01114083.24118103.941. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fsigned-char -ffunction-sections -fdata-sections -fno-strict-overflow -fwrapv
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUAVX-512 OffAVX-512 On20K40K60K80K100KMin: 50368.89 / Avg: 50413.98 / Max: 50489.9Min: 67626.19 / Avg: 67701.6 / Max: 67771.49Min: 73830.71 / Avg: 73906.76 / Max: 74055.95Min: 102946.01 / Avg: 103184.01 / Max: 103470.73Min: 112795.33 / Avg: 114083.24 / Max: 115070.32Min: 117003.98 / Avg: 118103.94 / Max: 120117.321. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fsigned-char -ffunction-sections -fdata-sections -fno-strict-overflow -fwrapv

Mobile Neural Network

MNN is the Mobile Neural Network as a highly efficient, lightweight deep learning framework developed by Alibaba. This MNN test profile is building the OpenMP / CPU threaded version for processor benchmarking and not any GPU-accelerated test. MNN does allow making use of AVX-512 extensions. Learn more via the OpenBenchmarking.org test page.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.1Model: SqueezeNetV1.0AVX-512 OffAVX-512 On3691215SE +/- 0.056, N = 10SE +/- 0.080, N = 15SE +/- 0.105, N = 3SE +/- 0.082, N = 3SE +/- 0.132, N = 3SE +/- 0.234, N = 94.1953.9995.2224.4159.2788.7621. (CXX) g++ options: -O3 -march=native -mno-amx-tile -std=c++11 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.1Model: SqueezeNetV1.0AVX-512 OffAVX-512 On3691215Min: 3.94 / Avg: 4.2 / Max: 4.38Min: 3.55 / Avg: 4 / Max: 4.34Min: 5.03 / Avg: 5.22 / Max: 5.39Min: 4.29 / Avg: 4.42 / Max: 4.57Min: 9.08 / Avg: 9.28 / Max: 9.53Min: 7.87 / Avg: 8.76 / Max: 9.611. (CXX) g++ options: -O3 -march=native -mno-amx-tile -std=c++11 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.1Model: mobilenetV3AVX-512 OffAVX-512 On0.8751.752.6253.54.375SE +/- 0.022, N = 10SE +/- 0.025, N = 15SE +/- 0.012, N = 3SE +/- 0.005, N = 3SE +/- 0.049, N = 3SE +/- 0.053, N = 91.7641.7092.2882.0553.8893.8661. (CXX) g++ options: -O3 -march=native -mno-amx-tile -std=c++11 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.1Model: mobilenetV3AVX-512 OffAVX-512 On246810Min: 1.68 / Avg: 1.76 / Max: 1.89Min: 1.59 / Avg: 1.71 / Max: 1.86Min: 2.27 / Avg: 2.29 / Max: 2.31Min: 2.05 / Avg: 2.06 / Max: 2.06Min: 3.82 / Avg: 3.89 / Max: 3.98Min: 3.59 / Avg: 3.87 / Max: 4.071. (CXX) g++ options: -O3 -march=native -mno-amx-tile -std=c++11 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

LeelaChessZero

LeelaChessZero (lc0 / lczero) is a chess engine automated vian neural networks. This test profile can be used for OpenCL, CUDA + cuDNN, and BLAS (CPU-based) benchmarking. Learn more via the OpenBenchmarking.org test page.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.28Backend: BLASAVX-512 OffAVX-512 On2K4K6K8K10KSE +/- 47.46, N = 5SE +/- 68.15, N = 3SE +/- 61.68, N = 3SE +/- 133.67, N = 3SE +/- 92.33, N = 9SE +/- 36.09, N = 347805524754210671942591711. (CXX) g++ options: -flto -O3 -march=native -mno-amx-tile -pthread
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgNodes Per Second Per Watt, More Is BetterLeelaChessZero 0.28Backend: BLASAVX-512 OffAVX-512 On61218243012.0914.0114.1119.8124.7024.19
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterLeelaChessZero 0.28CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 3292.51 / Max: 3407Min: 2300 / Avg: 3285.55 / Max: 3420Min: 1900 / Avg: 3429.01 / Max: 3524Min: 1900 / Avg: 3432.97 / Max: 3523Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709.02 / Max: 3731
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterLeelaChessZero 0.28CPU Power Consumption MonitorAVX-512 OffAVX-512 On130260390520650Min: 117.5 / Avg: 395.29 / Max: 593.12Min: 81.58 / Avg: 394.24 / Max: 583.5Min: 207.43 / Avg: 534.64 / Max: 715.83Min: 208.76 / Avg: 538.72 / Max: 765.81Min: 55.77 / Avg: 381.66 / Max: 520.36Min: 57.22 / Avg: 379.11 / Max: 511.65
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterLeelaChessZero 0.28CPU Temperature MonitorAVX-512 OffAVX-512 On1632486480Min: 45 / Avg: 60.47 / Max: 85Min: 54 / Avg: 71.37 / Max: 78Min: 36 / Avg: 48.76 / Max: 59Min: 35 / Avg: 44.16 / Max: 56Min: 33.63 / Avg: 46.72 / Max: 58.25Min: 33.25 / Avg: 40.84 / Max: 54.63
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.28Backend: BLASAVX-512 OffAVX-512 On2K4K6K8K10KMin: 4669 / Avg: 4780 / Max: 4922Min: 5428 / Avg: 5524.33 / Max: 5656Min: 7450 / Avg: 7541.67 / Max: 7659Min: 10441 / Avg: 10671 / Max: 10904Min: 9064 / Avg: 9425.11 / Max: 10021Min: 9108 / Avg: 9171.33 / Max: 92331. (CXX) g++ options: -flto -O3 -march=native -mno-amx-tile -pthread

Neural Magic DeepSparse

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On20406080100SE +/- 0.14, N = 3SE +/- 0.13, N = 3SE +/- 0.12, N = 3SE +/- 0.15, N = 3SE +/- 0.16, N = 3SE +/- 0.67, N = 345.8579.1446.8683.7071.51100.29
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On20406080100Min: 45.7 / Avg: 45.85 / Max: 46.13Min: 78.93 / Avg: 79.14 / Max: 79.37Min: 46.73 / Avg: 46.86 / Max: 47.1Min: 83.41 / Avg: 83.7 / Max: 83.86Min: 71.28 / Avg: 71.51 / Max: 71.83Min: 98.97 / Avg: 100.29 / Max: 101.19

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On510152025SE +/- 0.0648, N = 3SE +/- 0.0206, N = 3SE +/- 0.0534, N = 3SE +/- 0.0214, N = 3SE +/- 0.0319, N = 3SE +/- 0.0671, N = 321.797312.626421.321211.937813.97409.9667
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2936.97 / Max: 3405Min: 2300 / Avg: 2689.46 / Max: 3402Min: 1900 / Avg: 3338.2 / Max: 3518Min: 1900 / Avg: 3337.42 / Max: 3517Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On100200300400500Min: 122.5 / Avg: 322.04 / Max: 389.14Min: 83.01 / Avg: 313.16 / Max: 365.89Min: 110.3 / Avg: 466.39 / Max: 569.05Min: 214.75 / Avg: 471.13 / Max: 544.51Min: 30.08 / Avg: 340.51 / Max: 417.91Min: 61.86 / Avg: 340.27 / Max: 413.32
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 51 / Avg: 65.88 / Max: 73Min: 52 / Avg: 66.78 / Max: 74Min: 35 / Avg: 46.14 / Max: 52Min: 37 / Avg: 47.74 / Max: 53Min: 43 / Avg: 52.81 / Max: 58.25Min: 43 / Avg: 52.83 / Max: 57.75
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On510152025Min: 21.67 / Avg: 21.8 / Max: 21.87Min: 12.59 / Avg: 12.63 / Max: 12.66Min: 21.21 / Avg: 21.32 / Max: 21.38Min: 11.91 / Avg: 11.94 / Max: 11.98Min: 13.91 / Avg: 13.97 / Max: 14.02Min: 9.88 / Avg: 9.97 / Max: 10.1

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On1326395265SE +/- 0.09, N = 3SE +/- 0.10, N = 3SE +/- 0.05, N = 3SE +/- 0.17, N = 3SE +/- 0.08, N = 3SE +/- 0.02, N = 359.5249.4247.2137.2532.8328.98
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2834.61 / Max: 3407Min: 2300 / Avg: 2707.62 / Max: 3403Min: 1900 / Avg: 3338.32 / Max: 3520Min: 1900 / Avg: 3216.06 / Max: 3518Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On100200300400500Min: 116.53 / Avg: 308.23 / Max: 388.94Min: 116.85 / Avg: 307.49 / Max: 373.85Min: 209.04 / Avg: 468.14 / Max: 560.62Min: 206.44 / Avg: 463.48 / Max: 559.71Min: 58.11 / Avg: 334.36 / Max: 425.46Min: 56.93 / Avg: 328.16 / Max: 419.54
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 51 / Avg: 65.39 / Max: 73Min: 53 / Avg: 67.31 / Max: 75Min: 36 / Avg: 46.62 / Max: 53Min: 37 / Avg: 47.75 / Max: 53Min: 34.63 / Avg: 43.61 / Max: 48.38Min: 34.13 / Avg: 43.31 / Max: 48
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On1224364860Min: 59.35 / Avg: 59.52 / Max: 59.66Min: 49.27 / Avg: 49.42 / Max: 49.6Min: 47.14 / Avg: 47.21 / Max: 47.31Min: 36.93 / Avg: 37.25 / Max: 37.48Min: 32.72 / Avg: 32.83 / Max: 32.99Min: 28.95 / Avg: 28.98 / Max: 29.02

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On816243240SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.12, N = 3SE +/- 0.08, N = 3SE +/- 0.02, N = 316.8020.2321.1826.8430.4634.50
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On714212835Min: 16.76 / Avg: 16.8 / Max: 16.85Min: 20.16 / Avg: 20.23 / Max: 20.29Min: 21.13 / Avg: 21.18 / Max: 21.21Min: 26.68 / Avg: 26.84 / Max: 27.08Min: 30.31 / Avg: 30.46 / Max: 30.56Min: 34.45 / Avg: 34.5 / Max: 34.53

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On1326395265SE +/- 0.04, N = 3SE +/- 0.09, N = 3SE +/- 0.20, N = 3SE +/- 0.01, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 359.4249.5146.9736.9432.4029.20
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2813.09 / Max: 3401Min: 2300 / Avg: 2702.75 / Max: 3401Min: 1900 / Avg: 3376.77 / Max: 3517Min: 1900 / Avg: 3277.52 / Max: 3516Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On100200300400500Min: 118.87 / Avg: 314.48 / Max: 389.2Min: 115.91 / Avg: 309.32 / Max: 374.48Min: 203.35 / Avg: 468.13 / Max: 563.99Min: 106.45 / Avg: 468.4 / Max: 556.72Min: 42.78 / Avg: 344.31 / Max: 430.71Min: 60.73 / Avg: 336.08 / Max: 425.61
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 53 / Avg: 66.51 / Max: 75Min: 54 / Avg: 67.84 / Max: 75Min: 36 / Avg: 46.86 / Max: 52Min: 37 / Avg: 47.84 / Max: 53Min: 44.38 / Avg: 53.07 / Max: 59.13Min: 43.5 / Avg: 52.39 / Max: 57.75
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On1224364860Min: 59.36 / Avg: 59.42 / Max: 59.49Min: 49.34 / Avg: 49.51 / Max: 49.64Min: 46.58 / Avg: 46.97 / Max: 47.24Min: 36.92 / Avg: 36.94 / Max: 36.96Min: 32.32 / Avg: 32.4 / Max: 32.55Min: 29.08 / Avg: 29.2 / Max: 29.29

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On816243240SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.09, N = 3SE +/- 0.01, N = 3SE +/- 0.07, N = 3SE +/- 0.07, N = 316.8320.1921.2927.0730.8634.23
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On714212835Min: 16.81 / Avg: 16.83 / Max: 16.84Min: 20.14 / Avg: 20.19 / Max: 20.26Min: 21.17 / Avg: 21.29 / Max: 21.47Min: 27.05 / Avg: 27.07 / Max: 27.08Min: 30.71 / Avg: 30.86 / Max: 30.93Min: 34.13 / Avg: 34.23 / Max: 34.38

Intel Open Image Denoise

Open Image Denoise is a denoising library for ray-tracing and part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 1.4.0Run: RT.hdr_alb_nrm.3840x2160AVX-512 OffAVX-512 On0.94281.88562.82843.77124.714SE +/- 0.01, N = 4SE +/- 0.01, N = 5SE +/- 0.02, N = 5SE +/- 0.02, N = 6SE +/- 0.02, N = 5SE +/- 0.01, N = 52.072.942.954.193.463.48
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgImages / Sec Per Watt, More Is BetterIntel Open Image Denoise 1.4.0Run: RT.hdr_alb_nrm.3840x2160AVX-512 OffAVX-512 On0.00270.00540.00810.01080.01350.0060.0090.0070.0100.0120.012
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterIntel Open Image Denoise 1.4.0CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2644.39 / Max: 3400Min: 2300 / Avg: 2484.49 / Max: 3123Min: 1900 / Avg: 2976.52 / Max: 3510Min: 1900 / Avg: 2985.41 / Max: 3510Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterIntel Open Image Denoise 1.4.0CPU Power Consumption MonitorAVX-512 OffAVX-512 On130260390520650Min: 117.02 / Avg: 351.27 / Max: 568.33Min: 78.9 / Avg: 329.46 / Max: 563.85Min: 190.83 / Avg: 446.72 / Max: 747.36Min: 121.92 / Avg: 432.92 / Max: 757.14Min: 55.38 / Avg: 291.64 / Max: 587.02Min: 59.83 / Avg: 288.89 / Max: 584.26
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterIntel Open Image Denoise 1.4.0CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 50 / Avg: 61.2 / Max: 75Min: 48 / Avg: 59.52 / Max: 72Min: 37 / Avg: 43.04 / Max: 50Min: 37 / Avg: 42.14 / Max: 50Min: 35.88 / Avg: 40.59 / Max: 46.63Min: 35 / Avg: 39.27 / Max: 45.5
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 1.4.0Run: RT.hdr_alb_nrm.3840x2160AVX-512 OffAVX-512 On246810Min: 2.06 / Avg: 2.07 / Max: 2.08Min: 2.93 / Avg: 2.94 / Max: 2.96Min: 2.89 / Avg: 2.95 / Max: 2.99Min: 4.13 / Avg: 4.19 / Max: 4.24Min: 3.43 / Avg: 3.46 / Max: 3.52Min: 3.44 / Avg: 3.48 / Max: 3.51

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 1.4.0Run: RTLightmap.hdr.4096x4096AVX-512 OffAVX-512 On0.44780.89561.34341.79122.239SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 4SE +/- 0.00, N = 3SE +/- 0.00, N = 31.031.421.481.991.721.66
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgImages / Sec Per Watt, More Is BetterIntel Open Image Denoise 1.4.0Run: RTLightmap.hdr.4096x4096AVX-512 OffAVX-512 On0.00110.00220.00330.00440.00550.0030.0040.0030.0040.0050.005
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterIntel Open Image Denoise 1.4.0CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2632.24 / Max: 3400Min: 2300 / Avg: 2543.96 / Max: 3400Min: 1900 / Avg: 2888.03 / Max: 3505Min: 1900 / Avg: 2755.84 / Max: 3503Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterIntel Open Image Denoise 1.4.0CPU Power Consumption MonitorAVX-512 OffAVX-512 On130260390520650Min: 81.25 / Avg: 374.67 / Max: 565.97Min: 119.6 / Avg: 359.55 / Max: 567.51Min: 204.01 / Avg: 487.71 / Max: 755.13Min: 207.41 / Avg: 477.81 / Max: 753.94Min: 30.85 / Avg: 328.65 / Max: 583.24Min: 57.92 / Avg: 331.79 / Max: 581.42
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterIntel Open Image Denoise 1.4.0CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 47 / Avg: 59.15 / Max: 71Min: 45 / Avg: 57.82 / Max: 69Min: 35 / Avg: 43.41 / Max: 50Min: 35 / Avg: 42.32 / Max: 49Min: 34.63 / Avg: 40.87 / Max: 46.63Min: 33.25 / Avg: 39.68 / Max: 46.38
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 1.4.0Run: RTLightmap.hdr.4096x4096AVX-512 OffAVX-512 On246810Min: 1.03 / Avg: 1.03 / Max: 1.03Min: 1.42 / Avg: 1.42 / Max: 1.43Min: 1.48 / Avg: 1.48 / Max: 1.48Min: 1.97 / Avg: 1.99 / Max: 2Min: 1.72 / Avg: 1.72 / Max: 1.72Min: 1.65 / Avg: 1.66 / Max: 1.67

Neural Magic DeepSparse

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On60120180240300SE +/- 1.34, N = 3SE +/- 0.11, N = 3SE +/- 0.30, N = 3SE +/- 0.53, N = 3SE +/- 0.35, N = 3SE +/- 0.99, N = 13251.57180.17257.56168.23194.72135.99
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2856.54 / Max: 3402Min: 2300 / Avg: 2704.63 / Max: 3402Min: 1900 / Avg: 2956.37 / Max: 3516Min: 1900 / Avg: 2821.77 / Max: 3515Min: 3703 / Avg: 3708.97 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 124.13 / Avg: 437.31 / Max: 595.17Min: 103.25 / Avg: 453.02 / Max: 598.46Min: 211.06 / Avg: 598.84 / Max: 769.45Min: 184.38 / Avg: 608.53 / Max: 775.32Min: 32.05 / Avg: 494 / Max: 713.94Min: 31.38 / Avg: 537.92 / Max: 712.89
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 53 / Avg: 65.7 / Max: 74Min: 54 / Avg: 68.2 / Max: 75Min: 37 / Avg: 47.69 / Max: 53Min: 38 / Avg: 48.39 / Max: 53Min: 37.25 / Avg: 44.79 / Max: 51.25Min: 44.38 / Avg: 52.93 / Max: 60
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On50100150200250Min: 249.84 / Avg: 251.57 / Max: 254.22Min: 180.06 / Avg: 180.17 / Max: 180.38Min: 257.24 / Avg: 257.56 / Max: 258.16Min: 167.45 / Avg: 168.23 / Max: 169.24Min: 194.16 / Avg: 194.72 / Max: 195.37Min: 125.57 / Avg: 135.99 / Max: 140.35

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On1632486480SE +/- 0.18, N = 3SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 373.7547.8271.9340.1467.9048.70
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2701.11 / Max: 3401Min: 2300 / Avg: 2582.73 / Max: 3401Min: 1900 / Avg: 2690.42 / Max: 3512Min: 1900 / Avg: 2613.81 / Max: 3508Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 130.88 / Avg: 488.76 / Max: 594.17Min: 116.14 / Avg: 490.29 / Max: 592.28Min: 207.09 / Avg: 636.05 / Max: 765.98Min: 212.93 / Avg: 636.94 / Max: 774.46Min: 32.8 / Avg: 552.07 / Max: 653.8Min: 56.51 / Avg: 539.83 / Max: 646.46
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 57 / Avg: 73.09 / Max: 77Min: 54 / Avg: 71.83 / Max: 77Min: 37 / Avg: 49.9 / Max: 54Min: 37 / Avg: 50.14 / Max: 54Min: 43 / Avg: 56.4 / Max: 61.38Min: 42.13 / Avg: 55.39 / Max: 60.88
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On1428425670Min: 73.43 / Avg: 73.75 / Max: 74.07Min: 47.74 / Avg: 47.82 / Max: 47.93Min: 71.83 / Avg: 71.93 / Max: 72Min: 40.05 / Avg: 40.14 / Max: 40.19Min: 67.86 / Avg: 67.9 / Max: 67.94Min: 48.65 / Avg: 48.7 / Max: 48.8

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Detection,YOLOv5s COCO - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On306090120150SE +/- 0.21, N = 3SE +/- 0.06, N = 3SE +/- 0.59, N = 3SE +/- 1.49, N = 15SE +/- 1.97, N = 15SE +/- 2.25, N = 1569.7383.4284.84100.54109.26126.01
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Detection,YOLOv5s COCO - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On20406080100Min: 69.44 / Avg: 69.73 / Max: 70.13Min: 83.36 / Avg: 83.42 / Max: 83.53Min: 84.03 / Avg: 84.84 / Max: 85.99Min: 86.23 / Avg: 100.54 / Max: 106.06Min: 99.65 / Avg: 109.26 / Max: 119.75Min: 116.36 / Avg: 126.01 / Max: 135.4

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On2004006008001000SE +/- 2.25, N = 3SE +/- 0.79, N = 3SE +/- 2.38, N = 3SE +/- 0.72, N = 3SE +/- 0.15, N = 3SE +/- 1.43, N = 3913.42821.96889.39647.75787.721132.61
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2851.54 / Max: 3409Min: 2300 / Avg: 2747.12 / Max: 3411Min: 1900 / Avg: 2907.36 / Max: 3519Min: 1900 / Avg: 2733.68 / Max: 3518Min: 3709 / Avg: 3709 / Max: 3709Min: 3697 / Avg: 3708.93 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 120.76 / Avg: 424.25 / Max: 596.15Min: 116.32 / Avg: 423.52 / Max: 596.26Min: 179.48 / Avg: 575.96 / Max: 772.34Min: 184.97 / Avg: 576.22 / Max: 776.65Min: 45.38 / Avg: 492.05 / Max: 716.77Min: 62.09 / Avg: 449.53 / Max: 626.83
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 51 / Avg: 64.77 / Max: 74Min: 54 / Avg: 66.83 / Max: 75Min: 37 / Avg: 46.66 / Max: 54Min: 38 / Avg: 47.79 / Max: 53Min: 34.63 / Avg: 44.85 / Max: 55.13Min: 34.63 / Avg: 43.53 / Max: 49.5
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On2004006008001000Min: 909.13 / Avg: 913.42 / Max: 916.77Min: 820.58 / Avg: 821.96 / Max: 823.32Min: 886.1 / Avg: 889.39 / Max: 894.02Min: 646.33 / Avg: 647.75 / Max: 648.62Min: 787.55 / Avg: 787.72 / Max: 788.02Min: 1129.76 / Avg: 1132.61 / Max: 1134.13

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On2004006008001000SE +/- 1.29, N = 3SE +/- 3.51, N = 3SE +/- 1.38, N = 3SE +/- 0.76, N = 3SE +/- 0.46, N = 3SE +/- 2.80, N = 3916.94819.53887.59648.20787.331128.85
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2835.42 / Max: 3408Min: 2300 / Avg: 2731.23 / Max: 3408Min: 1900 / Avg: 2897.84 / Max: 3517Min: 1900 / Avg: 2758.89 / Max: 3517Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 121.91 / Avg: 423.91 / Max: 595.14Min: 109.29 / Avg: 424.09 / Max: 596.7Min: 205.49 / Avg: 579.26 / Max: 771.92Min: 111.58 / Avg: 575.15 / Max: 772.59Min: 60.66 / Avg: 500.64 / Max: 721.31Min: 61.67 / Avg: 453.86 / Max: 632.57
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 51 / Avg: 64.63 / Max: 73Min: 54 / Avg: 67.15 / Max: 75Min: 37 / Avg: 46.87 / Max: 52Min: 38 / Avg: 47.81 / Max: 53Min: 43.88 / Avg: 54.3 / Max: 62.25Min: 43.5 / Avg: 52.19 / Max: 59.13
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On2004006008001000Min: 914.37 / Avg: 916.94 / Max: 918.44Min: 812.53 / Avg: 819.53 / Max: 823.43Min: 886.01 / Avg: 887.59 / Max: 890.35Min: 646.74 / Avg: 648.2 / Max: 649.31Min: 786.78 / Avg: 787.33 / Max: 788.25Min: 1123.96 / Avg: 1128.85 / Max: 1133.68

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgItems / Sec, More Is BetterOpenVKL 1.3.1Benchmark: vklBenchmark ISPCAVX-512 OffAVX-512 On30060090012001500SE +/- 1.86, N = 3SE +/- 1.76, N = 3SE +/- 4.16, N = 3SE +/- 2.52, N = 3SE +/- 13.87, N = 3SE +/- 16.33, N = 37929171034119611761341
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgItems / Sec Per Watt, More Is BetterOpenVKL 1.3.1Benchmark: vklBenchmark ISPCAVX-512 OffAVX-512 On0.64671.29341.94012.58683.23351.7342.0141.7262.0572.4052.874
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOpenVKL 1.3.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2988.69 / Max: 3410Min: 2262 / Avg: 2929.05 / Max: 3409Min: 1899 / Avg: 2931.06 / Max: 3519Min: 1169 / Avg: 2887.71 / Max: 3515Min: 2621 / Avg: 3442.04 / Max: 3709Min: 2591 / Avg: 3486.01 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOpenVKL 1.3.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 101.07 / Avg: 456.78 / Max: 652.79Min: 100.02 / Avg: 455.32 / Max: 653.28Min: 183.35 / Avg: 599.21 / Max: 804.94Min: 179.36 / Avg: 581.38 / Max: 785.63Min: 60.18 / Avg: 488.95 / Max: 748.42Min: 57.24 / Avg: 466.56 / Max: 706.65
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOpenVKL 1.3.1CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 50 / Avg: 69.77 / Max: 80Min: 46 / Avg: 69.3 / Max: 78Min: 36 / Avg: 49.72 / Max: 57Min: 35 / Avg: 48.73 / Max: 56Min: 35.5 / Avg: 47.03 / Max: 54.63Min: 34.13 / Avg: 45.24 / Max: 52.88
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgItems / Sec, More Is BetterOpenVKL 1.3.1Benchmark: vklBenchmark ISPCAVX-512 OffAVX-512 On2004006008001000Min: 790 / Avg: 792.33 / Max: 796Min: 914 / Avg: 917.33 / Max: 920Min: 1028 / Avg: 1034 / Max: 1042Min: 1191 / Avg: 1196 / Max: 1199Min: 1151 / Avg: 1175.67 / Max: 1199Min: 1311 / Avg: 1341.33 / Max: 1367

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUAVX-512 OffAVX-512 On0.35550.7111.06651.4221.7775SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 31.581.161.571.041.140.961. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fsigned-char -ffunction-sections -fdata-sections -fno-strict-overflow -fwrapv
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2022.3CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2585.83 / Max: 3400Min: 2300 / Avg: 2378.88 / Max: 3400Min: 1900 / Avg: 2584.6 / Max: 3508Min: 1900 / Avg: 2332.5 / Max: 3508Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 120.36 / Avg: 516.2 / Max: 593.64Min: 119.8 / Avg: 517.02 / Max: 593.95Min: 225.78 / Avg: 674.16 / Max: 768.69Min: 211.51 / Avg: 670.46 / Max: 768.59Min: 30.32 / Avg: 579.7 / Max: 638.45Min: 61.47 / Avg: 583.01 / Max: 643.67
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterOpenVINO 2022.3CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 54 / Avg: 73.27 / Max: 77Min: 54 / Avg: 73.47 / Max: 77Min: 40 / Avg: 53.81 / Max: 57Min: 39 / Avg: 52.99 / Max: 56Min: 36.75 / Avg: 50.69 / Max: 53.75Min: 36.38 / Avg: 50.87 / Max: 54.63
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUAVX-512 OffAVX-512 On246810Min: 1.57 / Avg: 1.58 / Max: 1.58Min: 1.16 / Avg: 1.16 / Max: 1.16Min: 1.57 / Avg: 1.57 / Max: 1.57Min: 1.04 / Avg: 1.04 / Max: 1.04Min: 1.13 / Avg: 1.14 / Max: 1.14Min: 0.96 / Avg: 0.96 / Max: 0.971. (CXX) g++ options: -O3 -march=native -mno-amx-tile -fsigned-char -ffunction-sections -fdata-sections -fno-strict-overflow -fwrapv

Neural Magic DeepSparse

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On50100150200250SE +/- 0.69, N = 3SE +/- 0.35, N = 3SE +/- 0.84, N = 3SE +/- 0.32, N = 3SE +/- 0.10, N = 3SE +/- 0.09, N = 3217.10175.69226.27141.14187.35155.91
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2976.98 / Max: 3411Min: 2300 / Avg: 2795.9 / Max: 3409Min: 1900 / Avg: 3038.98 / Max: 3517Min: 1900 / Avg: 2810.29 / Max: 3515Min: 3709 / Avg: 3709 / Max: 3709Min: 3700 / Avg: 3708.95 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 79.98 / Avg: 423.53 / Max: 591.06Min: 116.32 / Avg: 423.16 / Max: 593.15Min: 202.66 / Avg: 577.34 / Max: 773.65Min: 213.45 / Avg: 574.76 / Max: 771.15Min: 58.55 / Avg: 490.27 / Max: 722.41Min: 58.49 / Avg: 475.62 / Max: 691.71
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 39 / Avg: 62.34 / Max: 77Min: 37 / Avg: 61.39 / Max: 74Min: 36 / Avg: 46.12 / Max: 53Min: 37 / Avg: 47.32 / Max: 53Min: 29.25 / Avg: 48.39 / Max: 60.5Min: 31.88 / Avg: 48.49 / Max: 59.25
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On4080120160200Min: 215.84 / Avg: 217.1 / Max: 218.23Min: 175.06 / Avg: 175.69 / Max: 176.27Min: 225.13 / Avg: 226.27 / Max: 227.9Min: 140.74 / Avg: 141.14 / Max: 141.77Min: 187.23 / Avg: 187.35 / Max: 187.56Min: 155.76 / Avg: 155.91 / Max: 156.08

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: CV Detection,YOLOv5s COCO - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On4080120160200SE +/- 0.40, N = 3SE +/- 0.17, N = 3SE +/- 0.14, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.36, N = 3161.33125.16147.46101.04144.89118.35
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2883.8 / Max: 3400Min: 2300 / Avg: 2672.86 / Max: 3410Min: 1900 / Avg: 2540.23 / Max: 3513Min: 1900 / Avg: 2474.62 / Max: 3509Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 121.41 / Avg: 485.92 / Max: 594.32Min: 117.09 / Avg: 488.17 / Max: 595.98Min: 212.95 / Avg: 636.84 / Max: 773.77Min: 108.87 / Avg: 636.97 / Max: 776.52Min: 33.52 / Avg: 555.47 / Max: 671.66Min: 59.96 / Avg: 563.77 / Max: 673.82
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 51 / Avg: 70.06 / Max: 77Min: 51 / Avg: 70.25 / Max: 76Min: 35 / Avg: 49.22 / Max: 53Min: 36 / Avg: 49.76 / Max: 54Min: 33.63 / Avg: 46.39 / Max: 50.63Min: 33.63 / Avg: 46.32 / Max: 50.63
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: CV Detection,YOLOv5s COCO - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On306090120150Min: 160.54 / Avg: 161.33 / Max: 161.85Min: 124.99 / Avg: 125.16 / Max: 125.5Min: 147.22 / Avg: 147.46 / Max: 147.7Min: 100.95 / Avg: 101.04 / Max: 101.1Min: 144.79 / Avg: 144.89 / Max: 144.98Min: 117.97 / Avg: 118.35 / Max: 119.07

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On20406080100SE +/- 0.17, N = 3SE +/- 0.18, N = 3SE +/- 0.06, N = 3SE +/- 0.18, N = 3SE +/- 0.04, N = 3SE +/- 0.26, N = 3105.2687.87102.6168.4395.1779.50
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 2912.11 / Max: 3406Min: 2300 / Avg: 2658.36 / Max: 3406Min: 1900 / Avg: 2946.44 / Max: 3517Min: 1900 / Avg: 2739.54 / Max: 3513Min: 3709 / Avg: 3709 / Max: 3709Min: 3709 / Avg: 3709 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On140280420560700Min: 122.8 / Avg: 469.84 / Max: 592.57Min: 101.9 / Avg: 468.91 / Max: 551.9Min: 203.11 / Avg: 619.84 / Max: 770.48Min: 211.14 / Avg: 618.61 / Max: 769.39Min: 60.72 / Avg: 532.23 / Max: 674.52Min: 31.6 / Avg: 498.87 / Max: 641.77
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Temperature MonitorAVX-512 OffAVX-512 On1530456075Min: 54 / Avg: 70.96 / Max: 78Min: 53 / Avg: 67.48 / Max: 74Min: 36 / Avg: 49.25 / Max: 54Min: 38 / Avg: 48.66 / Max: 53Min: 42.63 / Avg: 55.22 / Max: 61Min: 41.25 / Avg: 53.22 / Max: 60
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-StreamAVX-512 OffAVX-512 On20406080100Min: 104.94 / Avg: 105.26 / Max: 105.5Min: 87.58 / Avg: 87.87 / Max: 88.21Min: 102.51 / Avg: 102.61 / Max: 102.71Min: 68.14 / Avg: 68.43 / Max: 68.75Min: 95.09 / Avg: 95.17 / Max: 95.24Min: 78.98 / Avg: 79.5 / Max: 79.84

SMHasher

SMHasher is a hash function tester supporting various algorithms and able to make use of AVX and other modern CPU instruction set extensions. Learn more via the OpenBenchmarking.org test page.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMiB/sec, More Is BetterSMHasher 2022-08-22Hash: FarmHash32 x86_64 AVXAVX-512 OffAVX-512 On9K18K27K36K45KSE +/- 181.36, N = 4SE +/- 1.95, N = 4SE +/- 181.17, N = 4SE +/- 68.33, N = 4SE +/- 41.48, N = 5SE +/- 0.56, N = 530330.5333979.9937809.7142508.3536458.3139795.361. (CXX) g++ options: -O3 -march=native -mno-amx-tile -flto=auto -fno-fat-lto-objects
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgcycles/hash, Fewer Is BetterSMHasher 2022-08-22Hash: FarmHash32 x86_64 AVXAVX-512 OffAVX-512 On714212835SE +/- 0.01, N = 4SE +/- 0.02, N = 4SE +/- 0.00, N = 4SE +/- 0.00, N = 4SE +/- 0.00, N = 5SE +/- 0.16, N = 531.5831.0823.2723.2726.4126.551. (CXX) g++ options: -O3 -march=native -mno-amx-tile -flto=auto -fno-fat-lto-objects
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMiB/sec, More Is BetterSMHasher 2022-08-22Hash: FarmHash32 x86_64 AVXAVX-512 OffAVX-512 On7K14K21K28K35KMin: 29795.81 / Avg: 30330.53 / Max: 30569.12Min: 33974.33 / Avg: 33979.99 / Max: 33983.28Min: 37290.88 / Avg: 37809.71 / Max: 38097.38Min: 42303.42 / Avg: 42508.35 / Max: 42580.42Min: 36299.38 / Avg: 36458.31 / Max: 36530.98Min: 39793.53 / Avg: 39795.36 / Max: 39796.841. (CXX) g++ options: -O3 -march=native -mno-amx-tile -flto=auto -fno-fat-lto-objects

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.

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Large Mesh Size - Mesh TimeAVX-512 OffAVX-512 On2004006008001000887.211051.351039.32847.91867.38996.781. (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

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Medium Mesh Size - Mesh TimeAVX-512 OffAVX-512 On4080120160200162.72160.66169.84159.67146.75145.901. (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

CPU Power Consumption Monitor

Xeon 8380Xeon 8490HOpenBenchmarking.orgWattsCPU Power Consumption MonitorPhoronix Test Suite System MonitoringAVX-512 OffAVX-512 On160320480640800Min: 57.8 / Avg: 365.97 / Max: 660.32Min: 60.6 / Avg: 451.19 / Max: 668.12Min: 101.7 / Avg: 588.67 / Max: 887.42Min: 102.55 / Avg: 582.55 / Max: 894.12

CPU Peak Freq (Highest CPU Core Frequency) Monitor

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertzCPU Peak Freq (Highest CPU Core Frequency) MonitorPhoronix Test Suite System MonitoringAVX-512 OffAVX-512 On7001400210028003500Min: 802 / Avg: 3114.77 / Max: 3421Min: 804 / Avg: 2939.27 / Max: 3420Min: 1899 / Avg: 3041.53 / Max: 4080Min: 690 / Avg: 3000.31 / Max: 4081Min: 2336 / Avg: 3667.6 / Max: 4199Min: 2547 / Avg: 3615.9 / Max: 3772

Neural Magic DeepSparse

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 51 / Avg: 66.11 / Max: 71Min: 54 / Avg: 65.21 / Max: 69Min: 36 / Avg: 45.53 / Max: 49Min: 38 / Avg: 45.45 / Max: 49Min: 34.13 / Avg: 42.84 / Max: 47.75Min: 36.38 / Avg: 42.68 / Max: 48

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On90180270360450Min: 120.42 / Avg: 314.45 / Max: 351.47Min: 114.08 / Avg: 296.11 / Max: 324.93Min: 207.58 / Avg: 466.79 / Max: 528.2Min: 110.94 / Avg: 442.49 / Max: 492.85Min: 54.8 / Avg: 320.85 / Max: 377.94Min: 32.99 / Avg: 316.15 / Max: 367.34

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 3086.6 / Max: 3400Min: 2300 / Avg: 2877.68 / Max: 3400Min: 1900 / Avg: 3335.14 / Max: 3516Min: 1900 / Avg: 3282.46 / Max: 3518Min: 3697 / Avg: 3708.98 / Max: 3709Min: 3700 / Avg: 3708.99 / Max: 3709

Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: CV Detection,YOLOv5s COCO - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On48121620SE +/- 0.0435, N = 3SE +/- 0.0082, N = 3SE +/- 0.0817, N = 3SE +/- 0.1607, N = 15SE +/- 0.1629, N = 15SE +/- 0.1413, N = 1514.328611.976911.78239.97349.18357.9604
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgMegahertz, More Is BetterNeural Magic DeepSparse 1.1CPU Peak Freq (Highest CPU Core Frequency) MonitorAVX-512 OffAVX-512 On6001200180024003000Min: 2300 / Avg: 3086.6 / Max: 3400Min: 2300 / Avg: 2877.68 / Max: 3400Min: 1900 / Avg: 3335.14 / Max: 3516Min: 1900 / Avg: 3282.46 / Max: 3518Min: 3697 / Avg: 3708.98 / Max: 3709Min: 3700 / Avg: 3708.99 / Max: 3709
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Power Consumption MonitorAVX-512 OffAVX-512 On90180270360450Min: 120.42 / Avg: 314.45 / Max: 351.47Min: 114.08 / Avg: 296.11 / Max: 324.93Min: 207.58 / Avg: 466.79 / Max: 528.2Min: 110.94 / Avg: 442.49 / Max: 492.85Min: 54.8 / Avg: 320.85 / Max: 377.94Min: 32.99 / Avg: 316.15 / Max: 367.34
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgCelsius, Fewer Is BetterNeural Magic DeepSparse 1.1CPU Temperature MonitorAVX-512 OffAVX-512 On1428425670Min: 51 / Avg: 66.11 / Max: 71Min: 54 / Avg: 65.21 / Max: 69Min: 36 / Avg: 45.53 / Max: 49Min: 38 / Avg: 45.45 / Max: 49Min: 34.13 / Avg: 42.84 / Max: 47.75Min: 36.38 / Avg: 42.68 / Max: 48
Xeon 8380Xeon 8490HEPYC 9654OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: CV Detection,YOLOv5s COCO - Scenario: Synchronous Single-StreamAVX-512 OffAVX-512 On48121620Min: 14.24 / Avg: 14.33 / Max: 14.39Min: 11.96 / Avg: 11.98 / Max: 11.99Min: 11.62 / Avg: 11.78 / Max: 11.9Min: 9.42 / Avg: 9.97 / Max: 11.59Min: 8.34 / Avg: 9.18 / Max: 10.02Min: 7.37 / Avg: 7.96 / Max: 8.58

72 Results Shown

OpenVINO:
  Vehicle Detection FP16 - CPU
  Vehicle Detection FP16-INT8 - CPU
Cpuminer-Opt:
  LBC, LBRY Credits
  Quad SHA-256, Pyrite
OpenVINO
Cpuminer-Opt:
  Triple SHA-256, Onecoin
  Skeincoin
Neural Magic DeepSparse
miniBUDE:
  OpenMP - BM2:
    GFInst/s
    Billion Interactions/s
Cpuminer-Opt:
  scrypt
  x25x
miniBUDE:
  OpenMP - BM1:
    Billion Interactions/s
    GFInst/s
Neural Magic DeepSparse:
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream
  NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Stream
Mobile Neural Network
Neural Magic DeepSparse
OpenFOAM
Neural Magic DeepSparse
OSPRay:
  gravity_spheres_volume/dim_512/scivis/real_time
  gravity_spheres_volume/dim_512/ao/real_time
OpenVINO
Mobile Neural Network
Embree
Cpuminer-Opt
OSPRay Studio
Neural Magic DeepSparse:
  NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream
  NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream
oneDNN
OSPRay
OSPRay Studio:
  1 - 4K - 32 - Path Tracer
  1 - 4K - 16 - Path Tracer
  3 - 4K - 16 - Path Tracer
  1 - 4K - 1 - Path Tracer
  3 - 4K - 1 - Path Tracer
OpenFOAM
Neural Magic DeepSparse:
  CV Classification, ResNet-50 ImageNet - Synchronous Single-Stream:
    ms/batch
    items/sec
Embree
LeelaChessZero
OpenVINO
Mobile Neural Network:
  SqueezeNetV1.0
  mobilenetV3
LeelaChessZero
Neural Magic DeepSparse:
  NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Synchronous Single-Stream:
    items/sec
    ms/batch
  NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Stream:
    ms/batch
    items/sec
  NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Stream:
    ms/batch
    items/sec
Intel Open Image Denoise:
  RT.hdr_alb_nrm.3840x2160
  RTLightmap.hdr.4096x4096
Neural Magic DeepSparse:
  NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Stream
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream
  CV Detection,YOLOv5s COCO - Synchronous Single-Stream
  NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream
  NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream
OpenVKL
OpenVINO
Neural Magic DeepSparse:
  NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Stream
  CV Detection,YOLOv5s COCO - Asynchronous Multi-Stream
  NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream
SMHasher
OpenFOAM:
  drivaerFastback, Large Mesh Size - Mesh Time
  drivaerFastback, Medium Mesh Size - Mesh Time
CPU Power Consumption Monitor:
  Phoronix Test Suite System Monitoring:
    Watts
    Megahertz
  CPU Temp Monitor:
    Celsius
  CPU Power Consumption Monitor:
    Watts
  CPU Peak Freq (Highest CPU Core Frequency) Monitor:
    Megahertz
Neural Magic DeepSparse

Xeon 8380: AVX-512 Off

Kernel Notes: Transparent Huge Pages: madvise
Environment Notes: CXXFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile" CFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile"
Compiler 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-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-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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 -v
Processor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd000375
Python Notes: Python 3.10.7
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + retbleed: 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 IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 13 January 2023 21:21 by user phoronix.

Xeon 8380: AVX-512 On

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 SAMSUNG MZWLJ1T9HBJR-00007, Graphics: ASPEED, Monitor: VGA HDMI, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP

OS: Ubuntu 22.10, Kernel: 5.19.0-29-generic (x86_64), Desktop: GNOME Shell 43.1, Display Server: X Server 1.21.1.4, Vulkan: 1.3.224, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1920x1080

Kernel Notes: Transparent Huge Pages: madvise
Environment Notes: CXXFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile" CFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile"
Compiler 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-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-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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 -v
Processor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd000375
Python Notes: Python 3.10.7
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + retbleed: 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 IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 13 January 2023 09:11 by user phoronix.

Xeon 8490H: AVX-512 Off

Kernel Notes: Transparent Huge Pages: madvise
Environment Notes: CXXFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile" CFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile"
Compiler 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-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-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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 -v
Processor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b0000c0
Python Notes: Python 3.10.7
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 12 January 2023 19:07 by user phoronix.

Xeon 8490H: AVX-512 On

Processor: 2 x Intel Xeon Platinum 8490H @ 3.50GHz (120 Cores / 240 Threads), Motherboard: Quanta Cloud S6Q-MB-MPS (3A10.uh BIOS), Chipset: Intel Device 1bce, Memory: 1008GB, Disk: 2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007, Graphics: ASPEED, Monitor: VGA HDMI, Network: 4 x Intel E810-C for QSFP + 2 x Intel X710 for 10GBASE-T

OS: Ubuntu 22.10, Kernel: 5.19.0-29-generic (x86_64), Desktop: GNOME Shell 43.1, Display Server: X Server 1.21.1.4, Vulkan: 1.3.224, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1920x1080

Kernel Notes: Transparent Huge Pages: madvise
Environment Notes: CXXFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile" CFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile"
Compiler 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-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-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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 -v
Processor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b0000c0
Python Notes: Python 3.10.7
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 11 January 2023 12:31 by user phoronix.

EPYC 9654: AVX-512 Off

Kernel Notes: Transparent Huge Pages: madvise
Environment Notes: CXXFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile" CFLAGS="-O3 -march=native -mno-avx512f -mno-amx-tile"
Compiler 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-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-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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 -v
Processor Notes: Scaling Governor: amd-pstate performance (Boost: Enabled) - CPU Microcode: 0xa10110d
Python Notes: Python 3.10.7
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 15 January 2023 15:34 by user phoronix.

EPYC 9654: AVX-512 On

Processor: 2 x AMD EPYC 9654 96-Core @ 3.71GHz (192 Cores / 384 Threads), Motherboard: AMD Titanite_4G (RTI1002E BIOS), Chipset: AMD Device 14a4, Memory: 1520GB, Disk: 2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007, Graphics: ASPEED, Monitor: VGA HDMI, Network: Broadcom NetXtreme BCM5720 PCIe

OS: Ubuntu 22.10, Kernel: 5.19.0-29-generic (x86_64), Desktop: GNOME Shell 43.1, Display Server: X Server 1.21.1.4, Vulkan: 1.3.224, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1920x1080

Kernel Notes: Transparent Huge Pages: madvise
Environment Notes: CXXFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile" CFLAGS="-O3 -march=native -mavx512f -mavx512cd -mavx512vl -mavx512bw -mavx512dq -mavx512ifma -mavx512vbmi -mprefer-vector-width=512 -mno-amx-tile"
Compiler 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-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-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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 -v
Processor Notes: Scaling Governor: amd-pstate performance (Boost: Enabled) - CPU Microcode: 0xa10110d
Python Notes: Python 3.10.7
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 16 January 2023 16:33 by user phoronix.