Linux 6.9 + Ubuntu 22.04 Intel AMD Server CPU Benchmarks

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 2405207-NE-LINUX69SE55
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
Show Result Confidence Charts

Limit displaying results to tests within:

AV1 2 Tests
BLAS (Basic Linear Algebra Sub-Routine) Tests 4 Tests
C++ Boost Tests 5 Tests
Chess Test Suite 2 Tests
Timed Code Compilation 8 Tests
C/C++ Compiler Tests 16 Tests
Compression Tests 3 Tests
CPU Massive 25 Tests
Creator Workloads 23 Tests
Cryptography 5 Tests
Database Test Suite 5 Tests
Encoding 7 Tests
Fortran Tests 7 Tests
Game Development 5 Tests
HPC - High Performance Computing 29 Tests
Imaging 3 Tests
Common Kernel Benchmarks 3 Tests
LAPACK (Linear Algebra Pack) Tests 3 Tests
Linear Algebra 2 Tests
Machine Learning 7 Tests
Molecular Dynamics 10 Tests
MPI Benchmarks 10 Tests
Multi-Core 44 Tests
NVIDIA GPU Compute 3 Tests
Intel oneAPI 7 Tests
OpenMPI Tests 18 Tests
Programmer / Developer System Benchmarks 12 Tests
Python 2 Tests
Raytracing 2 Tests
Renderers 5 Tests
Scientific Computing 15 Tests
Software Defined Radio 2 Tests
Server 7 Tests
Server CPU Tests 17 Tests
Single-Threaded 2 Tests
Video Encoding 7 Tests
Common Workstation Benchmarks 3 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

Additional Graphs

Show Perf Per Core/Thread Calculation Graphs Where Applicable
Show Perf Per Clock Calculation Graphs Where Applicable
Show Perf Per RAM Channel Calculation Graphs Where Applicable

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs
Condense Test Profiles With Multiple Version Results Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
EPYC 8534P
May 18
  1 Day, 2 Hours, 46 Minutes
EPYC 8534PN
May 19
  1 Day, 3 Hours, 27 Minutes
EPYC 9654
May 13
  1 Day, 1 Hour, 28 Minutes
EPYC 9654 2P
May 14
  1 Day, 5 Hours, 21 Minutes
EPYC 9684X
May 12
  23 Hours, 34 Minutes
EPYC 9684X 2P
May 11
  1 Day, 3 Hours, 52 Minutes
EPYC 9754
May 15
  1 Day, 5 Hours, 27 Minutes
EPYC 9754 2P
May 17
  1 Day, 8 Hours, 26 Minutes
Xeon Max 9468
April 21
  1 Day, 3 Hours, 35 Minutes
Xeon Max 9468 2P
April 22
  1 Day, 22 Minutes
Xeon Max 9480
April 26
  1 Day, 1 Hour, 29 Minutes
Xeon Max 9480 2P
April 27
  1 Day, 6 Hours, 2 Minutes
Xeon Platinum 8490H
April 16
  22 Hours, 13 Minutes
Xeon Platinum 8490H 2P
April 17
  22 Hours, 41 Minutes
Xeon Platinum 8592+
April 11
  22 Hours, 1 Minute
Xeon Platinum 8592+ 2P
April 08
  23 Hours, 31 Minutes
Invert Hiding All Results Option
  1 Day, 2 Hours, 16 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):


Linux 6.9 + Ubuntu 22.04 Intel AMD Server CPU BenchmarksProcessorMotherboardChipsetMemoryDiskGraphicsNetworkOSKernelCompilerFile-SystemScreen ResolutionEPYC 8534PEPYC 8534PNEPYC 9654EPYC 9654 2PEPYC 9684XEPYC 9684X 2PEPYC 9754EPYC 9754 2PXeon Max 9468Xeon Max 9468 2PXeon Max 9480Xeon Max 9480 2PXeon Platinum 8490HXeon Platinum 8490H 2PXeon Platinum 8592+Xeon Platinum 8592+ 2PAMD EPYC 8534P 64-Core @ 2.30GHz (64 Cores / 128 Threads)AMD Cinnabar (RCB1009C BIOS)AMD Device 14a46 x 32GB DDR5-4800MT/s Samsung M321R4GA0BB0-CQKMG2 x 1920GB KIOXIA KCD8XPUG1T92 + 2000GB Corsair MP700 PROASPEED2 x Broadcom NetXtreme BCM5720 PCIeUbuntu 24.046.9.0-060900rc3-generic (x86_64)GCC 13.2.0ext41920x1200AMD EPYC 8534PN 64-Core @ 2.00GHz (64 Cores / 128 Threads)2 x 1920GB KIOXIA KCD8XPUG1T92AMD EPYC 9654 96-Core @ 2.40GHz (96 Cores / 192 Threads)AMD Titanite_4G (RTI1007B BIOS)12 x 64GB DDR5-4800MT/s Samsung M321R8GA0BB0-CQKEGBroadcom NetXtreme BCM5720 PCIe2 x AMD EPYC 9654 96-Core @ 2.40GHz (192 Cores / 384 Threads)1520GBAMD EPYC 9684X 96-Core @ 2.55GHz (96 Cores / 192 Threads)12 x 64GB DDR5-4800MT/s Samsung M321R8GA0BB0-CQKEG2 x AMD EPYC 9684X 96-Core @ 2.55GHz (192 Cores / 384 Threads)1520GBAMD EPYC 9754 128-Core @ 2.25GHz (128 Cores / 256 Threads)12 x 64GB DDR5-4800MT/s Samsung M321R8GA0BB0-CQKEG2 x AMD EPYC 9754 128-Core @ 2.25GHz (256 Cores / 512 Threads)1520GBIntel Xeon Max 9468 @ 3.50GHz (48 Cores / 96 Threads)Quanta Cloud QuantaGrid D54Q-2U S6Q-MB-MPS (3B05.TEL4P1 BIOS)Intel Device 1bce576GB2 x Intel Xeon Max 9468 @ 3.50GHz (96 Cores / 192 Threads)1136GB2 x Intel X710 for 10GBASE-TIntel Xeon Max 9480 @ 3.50GHz (56 Cores / 112 Threads)576GB2 x Intel Xeon Max 9480 @ 3.50GHz (112 Cores / 224 Threads)1136GB2 x Intel X710 for 10GBASE-TIntel Xeon Platinum 8490H @ 3.50GHz (60 Cores / 120 Threads)512GB2 x Intel Xeon Platinum 8490H @ 3.50GHz (120 Cores / 240 Threads)1008GB2 x Intel X710 for 10GBASE-TINTEL XEON PLATINUM 8592+ @ 3.90GHz (64 Cores / 128 Threads)512GB2 x INTEL XEON PLATINUM 8592+ @ 3.90GHz (128 Cores / 256 Threads)1008GB2 x Intel X710 for 10GBASE-TOpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseCompiler 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-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-OiuXZC/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Details- EPYC 8534P: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa00213- EPYC 8534PN: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa00213- EPYC 9654: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa101144- EPYC 9654 2P: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa101144- EPYC 9684X: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa101144- EPYC 9684X 2P: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa101144- EPYC 9754: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa00116- EPYC 9754 2P: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa00116- Xeon Max 9468: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2c000290- Xeon Max 9468 2P: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2c000290- Xeon Max 9480: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2c000290- Xeon Max 9480 2P: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2c000290- Xeon Platinum 8490H: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b000590- Xeon Platinum 8490H 2P: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b000590- Xeon Platinum 8592+: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000200- Xeon Platinum 8592+ 2P: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000200Java Details- OpenJDK Runtime Environment (build 21.0.3-ea+7-Ubuntu-1build1)Python Details- Python 3.12.2Security Details- EPYC 8534P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected- EPYC 8534PN: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected- EPYC 9654: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected- EPYC 9654 2P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected- EPYC 9684X: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected- EPYC 9684X 2P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected- EPYC 9754: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected- EPYC 9754 2P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected- Xeon Max 9468: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected- Xeon Max 9468 2P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected- Xeon Max 9480: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected- Xeon Max 9480 2P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected- Xeon Platinum 8490H: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected- Xeon Platinum 8490H 2P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected- Xeon Platinum 8592+: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected- Xeon Platinum 8592+ 2P: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected

EPYC 8534PEPYC 8534PNEPYC 9654EPYC 9654 2PEPYC 9684XEPYC 9684X 2PEPYC 9754EPYC 9754 2PXeon Max 9468Xeon Max 9468 2PXeon Max 9480Xeon Max 9480 2PXeon Platinum 8490HXeon Platinum 8490H 2PXeon Platinum 8592+Xeon Platinum 8592+ 2PLogarithmic Result OverviewPhoronix Test SuiteBRL-CADNAMDXmrigStockfishJohn The RipperASKAPPrimesieveRELIONm-queensCoremarkOpenSSLSpeedbHelsingBlenderNAS Parallel BenchmarksPennantOSPRay StudioLlamafileXcompact3d Incompact3dPyTorchACES DGEMMLULESHRocksDBminiBUDE7-Zip CompressionAlgebraic Multi-Grid BenchmarkWRFGROMACSOSPRayOpenVKLEmbreePostgreSQLASTC EncoderminiFEHigh Performance Conjugate GradientIntel Open Image DenoiseoneDNNKvazaarOpenRadiossOpenVINOGPAWMemcachedONNX RuntimeTimed Node.js CompilationOpenFOAMParallel BZIP2 CompressionLAMMPS Molecular Dynamics SimulatorLuxCoreRenderTimed LLVM Compilationuvg266CloverLeafQuantLibY-CruncherGraphicsMagickTimed Linux Kernel CompilationGNU Octave BenchmarkTensorFlowClickHouseNWChemLiquid-DSPApache IoTDBTimed Godot Game Engine CompilationQMCPACKTimed FFmpeg CompilationTimed Gem5 CompilationTimed ImageMagick CompilationsrsRAN ProjectSVT-AV1x265PyBenchTimed Mesa CompilationNumpy Benchmarklibavif avifencZstd CompressionSecureMarkFFmpegGoogle SynthMarkWebP Image Encode

EPYC 8534PEPYC 8534PNEPYC 9654EPYC 9654 2PEPYC 9684XEPYC 9684X 2PEPYC 9754EPYC 9754 2PXeon Max 9468Xeon Max 9468 2PXeon Max 9480Xeon Max 9480 2PXeon Platinum 8490HXeon Platinum 8490H 2PXeon Platinum 8592+Xeon Platinum 8592+ 2PLogarithmic Per Watt Result OverviewPhoronix Test SuiteminiFEMemcachedWebP Image EncodePyTorchZstd CompressionSVT-AV1FFmpegsrsRAN ProjectNumpy Benchmarkx265Google SynthMarkminiBUDE7-Zip CompressionKvazaarSecureMarkuvg266LuxCoreRenderClickHouseACES DGEMMLiquid-DSPRocksDBSpeedbGraphicsMagickASKAPEmbreeOpenVKLQuantLibASTC EncoderTensorFlowJohn The RipperLAMMPS Molecular Dynamics SimulatorStockfishCoremarkNAS Parallel BenchmarksOSPRayXmrigHigh Performance Conjugate GradientOpenSSLGROMACSIntel Open Image DenoiseAlgebraic Multi-Grid BenchmarkLlamafileBRL-CADNAMDLULESHP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.M

Linux 6.9 + Ubuntu 22.04 Intel AMD Server CPU Benchmarksopenvino: Weld Porosity Detection FP16-INT8 - CPUbrl-cad: VGR Performance Metricaskap: tConvolve MPI - Degriddingaskap: tConvolve MPI - Griddingnpb: SP.Copenfoam: drivaerFastback, Medium Mesh Size - Execution Timeopenssl: SHA256pennant: sedovbigxmrig: GhostRider - 1Mopenssl: SHA512compress-7zip: Decompression Ratingjohn-the-ripper: Blowfishjohn-the-ripper: bcryptopenradioss: Chrysler Neon 1Mprimesieve: 1e13openvino: Person Detection FP16 - CPUllamafile: wizardcoder-python-34b-v1.0.Q6_K - CPUopenssl: ChaCha20incompact3d: input.i3d 193 Cells Per Directionopenssl: ChaCha20-Poly1305openvino: Weld Porosity Detection FP16-INT8 - CPUopenvino: Person Detection FP16 - CPUspeedb: Rand Readrocksdb: Rand Readopenvino: Person Re-Identification Retail FP16 - CPUopenvino: Road Segmentation ADAS FP16-INT8 - CPUjohn-the-ripper: WPA PSKopenssl: RSA4096relion: Basic - CPUopenvino: Vehicle Detection FP16-INT8 - CPUquantlib: Multi-Threadedm-queens: Time To Solvecoremark: CoreMark Size 666 - Iterations Per Secondblender: Classroom - CPU-Onlypgbench: 100 - 1000 - Read Only - Average Latencyprimesieve: 1e12pgbench: 100 - 1000 - Read Onlyopenssl: RSA4096onnx: ArcFace ResNet-100 - CPU - Parallelospray-studio: 1 - 4K - 16 - Path Tracer - CPUblender: Pabellon Barcelona - CPU-Onlyblender: Barbershop - CPU-Onlyospray-studio: 3 - 4K - 16 - Path Tracer - CPUpytorch: CPU - 256 - ResNet-50openssl: AES-256-GCMopenssl: AES-128-GCMopenvino: Road Segmentation ADAS FP16-INT8 - CPUnpb: LU.Copenvino: Face Detection Retail FP16-INT8 - CPUhelsing: 14 digitopenvino: Vehicle Detection FP16-INT8 - CPUpytorch: CPU - 64 - ResNet-50ospray-studio: 1 - 4K - 32 - Path Tracer - CPUpytorch: CPU - 512 - ResNet-50openvino: Handwritten English Recognition FP16-INT8 - CPUospray-studio: 3 - 4K - 32 - Path Tracer - CPUnpb: CG.Cnpb: IS.Dastcenc: Exhaustiveaskap: tConvolve MT - Griddingblender: BMW27 - CPU-Onlyblender: Fishy Cat - CPU-Onlyminibude: OpenMP - BM2minibude: OpenMP - BM2openvino: Person Vehicle Bike Detection FP16 - CPUastcenc: Very Thoroughmt-dgemm: Sustained Floating-Point Rateospray-studio: 1 - 4K - 1 - Path Tracer - CPUospray-studio: 3 - 4K - 1 - Path Tracer - CPUlulesh: speedb: Read While Writingpytorch: CPU - 256 - ResNet-152pytorch: CPU - 512 - ResNet-152openvino: Machine Translation EN To DE FP16 - CPUpytorch: CPU - 64 - ResNet-152npb: MG.Camg: srsran: PUSCH Processor Benchmark, Throughput Totalonnx: fcn-resnet101-11 - CPU - Parallelopenvino: Machine Translation EN To DE FP16 - CPUaircrack-ng: blender: Junkshop - CPU-Onlyospray: gravity_spheres_volume/dim_512/pathtracer/real_timeonnx: yolov4 - CPU - Parallelwrf: conus 2.5kmminibude: OpenMP - BM1minibude: OpenMP - BM1onnx: ArcFace ResNet-100 - CPU - Standardembree: Pathtracer ISPC - Crownonnx: T5 Encoder - CPU - Standardonnx: ResNet50 v1-12-int8 - CPU - Parallelgromacs: MPI CPU - water_GMX50_barekvazaar: Bosphorus 4K - Slowonnx: yolov4 - CPU - Standardllamafile: llava-v1.5-7b-q4 - CPUospray: particle_volume/ao/real_timeospray: particle_volume/scivis/real_timekvazaar: Bosphorus 4K - Mediumopenvkl: vklBenchmarkCPU ISPCincompact3d: X3D-benchmarking input.i3dopenvino: Person Vehicle Bike Detection FP16 - CPUliquid-dsp: 256 - 256 - 32embree: Pathtracer ISPC - Asian Dragon Objastcenc: Thoroughuvg266: Bosphorus 4K - Slowliquid-dsp: 256 - 256 - 512embree: Pathtracer ISPC - Asian Dragonhpcg: 144 144 144 - 60onnx: GPT-2 - CPU - Standardonnx: CaffeNet 12-int8 - CPU - Paralleloidn: RT.ldr_alb_nrm.3840x2160 - CPU-Onlyonnx: super-resolution-10 - CPU - Paralleloidn: RTLightmap.hdr.4096x4096 - CPU-Onlyopenvino: Noise Suppression Poconet-Like FP16 - CPUuvg266: Bosphorus 4K - Mediumoidn: RT.hdr_alb_nrm.3840x2160 - CPU-Onlygraphics-magick: Sharpenopenvino: Face Detection FP16-INT8 - CPUonednn: Deconvolution Batch shapes_3d - CPUonnx: CaffeNet 12-int8 - CPU - Standardliquid-dsp: 256 - 256 - 57onnx: ResNet50 v1-12-int8 - CPU - Standardluxcorerender: Orange Juice - CPUgpaw: Carbon Nanotubeopenvino: Noise Suppression Poconet-Like FP16 - CPUonnx: super-resolution-10 - CPU - Standardmemcached: 1:100openfoam: drivaerFastback, Small Mesh Size - Execution Timeopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUopenvino: Person Re-Identification Retail FP16 - CPUtensorflow: CPU - 512 - AlexNetgraphics-magick: Enhancedlammps: Rhodopsin Proteincloverleaf: clover_bm64_shortbuild-nodejs: Time To Compileonnx: GPT-2 - CPU - Parallelcompress-7zip: Compression Ratingonnx: bertsquad-12 - CPU - Parallelcompress-pbzip2: FreeBSD-13.0-RELEASE-amd64-memstick.img Compressionlammps: 20k Atomsopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUbuild-llvm: Ninjapgbench: 100 - 1000 - Read Write - Average Latencypgbench: 100 - 1000 - Read Writebuild-linux-kernel: allmodconfigonnx: bertsquad-12 - CPU - Standardopenradioss: INIVOL and Fluid Structure Interaction Drop Containercloverleaf: clover_bm16tensorflow: CPU - 256 - AlexNetastcenc: Mediumy-cruncher: 500Mtensorflow: CPU - 64 - GoogLeNetopenvino: Handwritten English Recognition FP16-INT8 - CPUy-cruncher: 1Bkvazaar: Bosphorus 4K - Very Fastliquid-dsp: 128 - 256 - 512openvino: Face Detection FP16-INT8 - CPUtensorflow: CPU - 512 - GoogLeNetluxcorerender: LuxCore Benchmark - CPUoctave-benchmark: graphics-magick: Swirlonnx: T5 Encoder - CPU - Parallelliquid-dsp: 128 - 256 - 32tensorflow: CPU - 256 - GoogLeNettensorflow: CPU - 512 - ResNet-50apache-iotdb: 800 - 100 - 800 - 100tensorflow: CPU - 64 - ResNet-50apache-iotdb: 800 - 100 - 800 - 100build-linux-kernel: defconfiguvg266: Bosphorus 4K - Very Fastliquid-dsp: 128 - 256 - 57clickhouse: 100M Rows Hits Dataset, Third Runclickhouse: 100M Rows Hits Dataset, Second Runnwchem: C240 Buckyballopenradioss: Cell Phone Drop Testclickhouse: 100M Rows Hits Dataset, First Run / Cold Cachetensorflow: CPU - 256 - ResNet-50liquid-dsp: 1 - 256 - 512build-godot: Time To Compileqmcpack: Li2_STO_aebuild-ffmpeg: Time To Compileapache-iotdb: 500 - 100 - 800 - 400openfoam: drivaerFastback, Medium Mesh Size - Mesh Timegraphics-magick: Noise-Gaussiantensorflow: CPU - 64 - AlexNetliquid-dsp: 1 - 256 - 57svt-av1: Preset 8 - Bosphorus 4Kbuild-gem5: Time To Compilebuild-imagemagick: Time To Compileapache-iotdb: 500 - 100 - 800 - 400liquid-dsp: 64 - 256 - 512appleseed: Disney Materialcompress-zstd: 19, Long Mode - Compression Speedopenvino: Face Detection Retail FP16-INT8 - CPUsvt-av1: Preset 12 - Bosphorus 4Kcompress-zstd: 19 - Decompression Speedx265: Bosphorus 4Ksvt-av1: Preset 4 - Bosphorus 4Kvvenc: Bosphorus 4K - Fasteropenfoam: drivaerFastback, Small Mesh Size - Mesh Timepybench: Total For Average Test Timesliquid-dsp: 64 - 256 - 57liquid-dsp: 64 - 256 - 32quantlib: Single-Threadedonnx: Faster R-CNN R-50-FPN-int8 - CPU - Parallelavifenc: 6srsran: PDSCH Processor Benchmark, Throughput Threadcompress-zstd: 19 - Compression Speedliquid-dsp: 1 - 256 - 32compress-zstd: 19, Long Mode - Decompression Speedbuild-mesa: Time To Compileavifenc: 6, Losslessnumpy: avifenc: 0avifenc: 10, Losslessavifenc: 2vvenc: Bosphorus 4K - Fastwebp: Quality 100, Lossless, Highest Compressionsecuremark: SecureMark-TLSffmpeg: libx265 - Uploadffmpeg: libx265 - Video On Demandsynthmark: VoiceMark_100ffmpeg: libx265 - Platformwebp: Quality 100, Highest Compressionsrsran: PUSCH Processor Benchmark, Throughput Threadstockfish: Chess Benchmarkonnx: T5 Encoder - CPU - Parallelonnx: T5 Encoder - CPU - Standardonnx: Faster R-CNN R-50-FPN-int8 - CPU - Parallelonnx: Faster R-CNN R-50-FPN-int8 - CPU - Standardonnx: Faster R-CNN R-50-FPN-int8 - CPU - Standardonnx: CaffeNet 12-int8 - CPU - Parallelonnx: CaffeNet 12-int8 - CPU - Standardonnx: ResNet50 v1-12-int8 - CPU - Parallelonnx: ResNet50 v1-12-int8 - CPU - Standardonnx: ArcFace ResNet-100 - CPU - Parallelonnx: ArcFace ResNet-100 - CPU - Standardonnx: GPT-2 - CPU - Parallelonnx: GPT-2 - CPU - Standardonnx: bertsquad-12 - CPU - Parallelonnx: bertsquad-12 - CPU - Standardonnx: super-resolution-10 - CPU - Parallelonnx: super-resolution-10 - CPU - Standardonnx: fcn-resnet101-11 - CPU - Parallelonnx: fcn-resnet101-11 - CPU - Standardonnx: fcn-resnet101-11 - CPU - Standardonnx: yolov4 - CPU - Parallelonnx: yolov4 - CPU - Standardsrsran: PDSCH Processor Benchmark, Throughput Totalsvt-av1: Preset 13 - Bosphorus 4Kluxcorerender: DLSC - CPUnpb: EP.Dpennant: leblancbignamd: STMV with 1,066,628 Atomsnamd: ATPase with 327,506 Atomsminife: Smallrocksdb: Read While WritingEPYC 8534PEPYC 8534PNEPYC 9654EPYC 9654 2PEPYC 9684XEPYC 9684X 2PEPYC 9754EPYC 9754 2PXeon Max 9468Xeon Max 9468 2PXeon Max 9480Xeon Max 9480 2PXeon Platinum 8490HXeon Platinum 8490H 2PXeon Platinum 8592+Xeon Platinum 8592+ 2P5366.7985199716704.416573.252745.64691.642247458472100711.369644660.024378460820367989102198102222297.2545.953161.485.1630706893837720.937913921118566172711.78197.912740917782661865373692.6731.99386212881480.8468.68810.78172566.813.2502274636.13154467.670.3843.801260650025894.627.31283881185.12240.134537733.68462309750530539784931373998.09133284.469256.11105.6762943.0833.497205133.761542.618426122229.942746.203.51465826.4427.1234.35107.7152692.8732830.125.742021.8175772043242116949.5911205390113.2613.28100.6713.1963930.6411500526673193.02.57278317.38153818.03135.8218.47058.7581416146.997107.3532683.83130.374362.1866127.085207.3366.49923.718.8744621.0017.966817.944824.311266567.28694711.22306220000064.325341.022517.0482101333375.214922.197289.2081547.5241.8097.05080.8617.3019.251.79112592.221.49835443.0583116933333219.05813.8163.7963636.33106.5266198256.7258.1692350.538.59875.2317224.24657.38277.525128.9823313579.80092.20049727.53287511.11205.01416.32161272406.41412.3884176.87495.25843.63294.62095.077276.9341.3810.20554.6080991666753.89313.507.388.929379255.6713047066667308.46101.1711687979987.3165.6644.27650.042982800000338.84337.322024.733.23323.5998.7011290000154.010137.2327.536105432862163.69864123673.424621066768.188242.93117.917251.4763386000056.4175186.766.77165.0891118.828.136.93913.96833.673162976218683333318713333332593.327.93623.484557.913.9298996671101.421.0357.182421.9975.8645.73541.6576.1180.4825002423.2546.88655.23946.873.03112.1991981443.910067.8732535.794327.928136.07721.824432.257164.825064.5702936.611932.92017.7452111.2047102.050380.723210.30269.38681388.743216.7034.66491114.174112.72819375.9160.7859.396416.346.5417681.519015.7067224526.093336454830.0177135215903.416399.752684.93690.735036780673038712.151324564.3221944709733383559347393472300.6954.389156.595.1128205332707721.178030719330509364313.11204.052429401102349926363355.7334.76339998810291.8496.24111.48153920.814.4932069432.52809375.260.4144.545241637521047.526.24754251394.41263.764884433.30415893963207488965390857918.53130839.128279.63120.5602763.4833.237809333.181396.789164321684.802718.943.19935821.4630.0738.0497.9782449.4542506.105.240620.9977322240264916855.8911188210313.1813.28109.0013.1562449.5511471080002818.52.25453293.07141831.77638.8616.92008.4728316609.78198.2072455.16331.064458.2776127.208203.0556.24622.468.7523320.6516.450416.450523.021177568.25404912.67278960000063.343337.959715.7774174000073.907322.134188.5688553.4711.6996.16790.8018.4917.801.69105654.831.79749446.6892793700000211.18713.0066.7053405.94102.9665820618.1859.2663770.69.46796.8216321.93557.54288.118129.6463190948.871442.32119525.33478334.92215.92518.05355395438.16811.5447187.03498.22771.47273.67035.165256.7945.7110.42751.5173175666748.73295.296.938.948363254.7272767933333287.4295.6311788497780.8664.9346.35244.922687066667335.46337.442224.635.95322.6992.8211236333159.624151.2728.420105138551188.37105120617.034548633364.787251.60618.175251.0557726333361.3624636.747.60157.3621144.827.746.72713.61436.823441974202633333317768666672593.027.87033.623563.613.8298993331096.521.5307.301418.8978.2415.73442.6056.0000.4825016223.2746.93656.25046.863.03112.1906581563.924557.8601535.883126.829037.53021.805732.238684.926534.7422138.102732.22947.7051111.2864112.76086.651510.39779.71131443.961226.6264.46533118.023114.31119485.8153.4178.485745.046.8527121.461635.4497124548.889341769356.71316121642172.446233.7125191.11325.423311264751376905.91645814207.739627575497642626166541166541176.1424.237157.458.355033110211238.2538351135096099835710.09306.274285570544092729496554.1227.466434751436324.9238.8818.58294183.27.7383926265.27712040.260.2782.058360326135964.929.26901889648.88139.252230534.907676965536678972884361231743.49269047.2816038.9062.1735528.1535.014329034.652739.015001953182.775229.415.630112394.615.9620.15171.7714294.2835024.329.184439.0862621180139126082.7111802499313.9514.0591.1613.84116725.0422828020005239.61.25135525.88186675.43221.0230.60739.7851610176.864136.9453423.63439.2586110.5253187.012190.38212.51438.909.8601825.7528.095628.068939.582077466.5480149.475271933333113.278365.302528.011362166667132.491828.3185126.816586.1782.73109.9981.3116.3231.442.73173510.101.15487604.0455266466667225.30821.3234.8665697.51122.6838220895.4532.5917830.487.231363.3326436.15146.50164.021156.61054950012.25491.42295442.507133779.62135.66612.50279987311.82513.8647100.24334.301278.58449.11413.850354.5934.926.74376.83115816666793.99470.9511.347.528511374.7454032733333447.61140.01142102254114.6353.7734.56762.024256300000559.55562.501575.522.53536.82138.1113394667111.688108.0220.494125562060136.39912160942.195542533387.993185.30013.471211.8581137333343.4540518.385.85178.5161388.931.728.42216.73028.031022818285983333322827666673105.929.83812.756684.717.2360683331328.516.9635.799529.8061.9924.79234.1487.4310.5729834127.6655.59783.77855.613.62112.21977083502.667115.3465633.520923.919842.06921.705791.660465.253214.4613534.163625.47896.377807.8858481.604172.12329.089558.15092800.231182.0175.55018102.191101.468627695.2172.87514.2910652.033.7548352.878778.1827550909.61460854418666.16616222881025.886757.7256584.67124.299462516456298873.40833726702.778822257383103327731649232052777.6412.63163.985.2410015026348604.2607859669892582222310.17750.1185768717981514511012024.1814.3012533332861959.8145.6474.46593246.94.6097975907.45064921.120.2931.242342273171942.015.00821042826.3579.051242717.41152650546581317850068790673353.73472931.0431974.8833.67510708.9117.072075016.245178.292794384477.7210139.6910.601812708.58.6111.25291.2787281.9589203.1917.009868.86850566178048760.815303245967.147.2048.437.04194942.3243249303336835.70.917043990.1612.8752.67416.348916918.021217.8465446.16123.5786176.2061167.107100.358620.23461.076.4122914.4747.614347.526761.113335248.8673155.207513333333175.5971105.405240.962129533333185.711042.5158102.572315.3553.4986.92301.6723.1242.033.48262253.631.009473444.8477411866667133.48529.6125.8447810.6595.83728498306.7825.0056320.353.981862.7936850.12130.05118.902109.4647617348.809301.12632957.099172688.5094.81018.05955505197.63010.661585.89266.101524.74491.53444.329227.2236.908.19672.681199966667188.74525.8410.567.645553271.6204063333333401.84161.9012559560782.9461.0727.67853.444280933333525.23529.801558.423.53504.84133.8713390667102.583102.9818.089108821580140.48261178727.775495466786.341164.89912.749240.438245833338.305.98171.2181382.128.818.34927.737712822273023333322820666673098.525.62652.810673.417.0360330001324.616.4095.846529.0362.8874.82935.4020.5729961026.4152.81782.32252.793.62112.23707515303.680015.9843639.017625.317139.80353.170642.250209.962457.5555866.628542.51379.124529.74614113.51793.834011.502410.43441094.256220.3514.53817157.505155.96022931.8167.22317.8919414.942.3129775.2783917.8110543531.0201082899997.89312257464260.274389.0225311.82178.253821384095889835.46066015786.64586513191371606519257419289877.3223.265110.918.485449040847736.881212793839256520779.42433.065376330345233100916846.7921.636713221613828.7176.4227.85319827.26.9504152864.87538035.660.2511.984398935848996.829.84931660344.87123.901953534.908597205999639960071092672212.41362511.4317230.3556.9126017.1934.713864534.692986.334434962071.106346.156.302714955.614.2918.36187.6844692.1065608.0710.322445.2457941037122527287.2611959577714.0813.9583.1113.89134994.1423761490005656.91.25179576.98206619.29218.6534.64479.800087885.16151.9483798.69541.4745130.7874279.641178.69815.22240.009.7368525.3832.518832.523240.502396309.8476538.485862466667133.476973.379629.181570400000155.120827.5824162.092590.7072.66108.8681.2813.5232.742.64201471.981.11923672.9165811900000243.34024.6930.0016963.31122.9488634258.3928.3255940.456.931392.9130039.99842.17147.837161.49757911212.28071.36906346.891138605.52124.45512.58079492290.34113.974594.76245.821301.26500.76734.312330.2632.007.47178.151368733333101.51425.2512.436.857546446.3034353533333406.58121.96144760448104.5252.9332.88962.704755333333595.19597.731450.520.45566.31123.3713387333103.35297.00419.991126161028118.54229165868.545545533386.298176.87612.199209.6885058666740.9189618.455.41165.7531385.332.128.59517.01625.931646818286200000022874666673114.228.72502.723680.717.2360660001324.916.6685.757542.1561.6294.77034.0447.5470.5730083327.6155.56783.42255.383.61112.22145246592.239223.5754334.812621.578446.87641.693071.485965.594914.1091433.499424.10976.184786.1666881.431971.55649.185248.13326801.961178.0325.66508102.035102.73429384.9162.47516.2111424.053.3604533.410859.3353453167.81415841920550.936193921123480121976.9390520.9393.5681152762677664033.08110831157.891455971200111646537458937626849.5512.13043.775.2410866181919503.753199017676527195479.251095.791059546533103120915112665.6211.9512940003214376.393.1684.08644091.04.1338448083.72998018.670.2891.199348090998237.015.0516891623.8467.521051416.97171406176108719903520599674013.96551955.5535809.5930.69411735.4217.281785516.825766.622089388487.9112296.2711.910714185.17.7510.16309.2197730.4719752.8919.405278.49615856166054678.485333075747.027.0743.177.08244738.8144672813336999.70.8576851110.4911.4760.37486.318284431.344236.6105915.25323.4960214.0178263.524100.460022.52562.827.2486914.2955.338255.257663.483906167.1281644.917935266667210.4367116.122643.392394333333231.427244.9502161.947310.5913.4787.25791.6422.2045.263.45297233.631.001914446.8058012700000141.83033.0824.1348231.72119.55858683410.7922.7079890.353.781897.6440853.77122.83114.912110.1687413668.759711.10944662.000178488.0389.27518.15955272181.99511.270779.42223.541560.56508.09264.510248.0333.188.99274.031394733333205.03464.7810.616.955583276.2074407900000391.40137.5012494728886.0861.4625.87255.054759366667526.36528.211429.922.52513.92120.111308313396.16191.96617.839106778379135.5163179738.315541566787.824161.93611.720246.088493400008.425.34163.7951392.531.958.51326.823627818277280000022793000003125.424.05402.890664.716.8360400001321.416.4015.950540.9663.4044.94335.3160.5630174027.4355.01783.20055.113.62112.23991242233.617923.8079241.569622.788644.06353.217812.246459.952457.0968266.435542.88799.066076.17190114.15890.062111.45808.399691165.93182.2685.49711158.265137.98925977.4161.33922.5822166.602.0768356.3834320.7525959466.61985166411508.19348447728476.628854.3127519.37449.465971612691622805.1476589320.552746634113780871220188220188231.9621.762230.247.476563963293878.8722591445703974887010.92283.185899510705732354527862.6538.218031871872096.1360.20710.97374413.56.3504525288.00301431.650.2921.921342249253966.924.60951528639.83114.761812927.7899753900787311672211418701671.23279397.4519917.5448.8495778.5127.683738327.533280.254294543799.785702.137.557312731.812.8616.58232.7495818.7365594.8112.318744.970826954113024218.8312282942611.0110.91102.8410.88125567.0722753526675973.83.13156621.42177989.89117.7037.73818.6359411895.407229.9135747.82330.8572126.1456131.004153.24613.24238.478.2953320.6435.246135.146439.082508534.05942811.356588200000133.929786.013929.001744700000156.390923.626689.7350452.3233.1694.73651.5022.6732.443.12207551.070.943621382.1036401366667160.90423.7633.1205310.07106.1307736018.7433.8142230.738.031620.5330440.38648.36186.103137.1935754409.652961.35573747.986132973.43146.76113.84372244337.68511.4182106.54400.521454.71527.79973.968314.6738.856.92868.561369800000115.78450.6112.418.801535319.4523791566667462.79135.17131781309106.1458.1138.52654.464714500000458.96450.12170023.69433.79125.7611174500128.279110.8023.217116685326167.40625151924.804627000078.739212.54616.388227.4070965000045.0780876.686.25169.2301139.928.167.28314.78530.381372974244493333318987666672596.127.44713.251596.213.8299073331102.219.9066.853425.4673.2715.66240.3516.4650.4825069523.3647.33655.79547.313.04112.11897637993.129047.6333736.446026.445837.95852.210202.616536.523676.2279140.633032.47447.2808711.1392103.59787.655210.55379.42177319.348166.7936.01164115.790120.55625194.6163.41316.7912575.303.6543032.888697.9975650139.91577618622499.03586175052639.554066.0214017.18137.113853202586862372.94045617873.1104776633047134022641541241272393.9911.55785.864.6313069285075734.5106189390602860899011.20750.961169947385113520656414601.5215.8415510003729470.8201.2655759922.13.9818171241.70851216.820.3291.3593044620107830.611.9382841321.6665.581001513.11198310302670023166119452834032.66588329.3238546.6626.77612750.5913.681696013.766400.942007163265.019771.4114.103811208.77.179.71288.5377213.43310182.1622.002384.68859453163046126.080365932045.715.7556.735.79245447.7227830390838396.72.517091127.0511.2163.84595.557677365.863214.6175365.42118.0133190.0839127.53788.374121.30554.886.7497710.9456.693656.448554.574012253.3185536.276439000000200.1612121.301737.131734800000208.313141.872786.6864270.6313.5778.06761.7128.1038.123.59310275.530.829653316.1136316266667117.61732.2627.1456513.3698.63747645325.7024.5701220.414.371876.6640351.62927.74147.73599.21827769687.984021.06839858.839166151.42101.56116.88459228210.19710.7518890.85349.671403.46462.23824.100176.0039.867.91559.011349800000231.83588.439.658.848567250.6343794133333397.44174.9311531783467.0866.5430.68046.004626566667427.95420.351712.627.86411.26131.3811273667126.112111.1521.224103630483192.41705164658.974622966777.596192.97715.810253.137093933336.646.6168.7791139.028.117.15329.474484972237476666718949666672590.826.32573.341576.813.7298933331104.219.5586.986422.7674.6895.74041.9160.4825107523.2847.14655.04647.133.03112.13012022323.987357.8406538.000027.897535.87123.696023.1644111.31488.5180883.766055.915810.066611.5774125.24493.636012.809310.23106397.429203.5714.91420179.989148.20829877.1159.74622.0022017.752.0843845.8360618.9745442113.52290015219183.837139371852221766.558815.36440.831134443918518721.021125446.8161992612772077386789367875228.7467.32441.931.4722659735415313.96923761613210582132.50285.952114570482053461546717.6555.74286552697707.2412.37116.33143755.321.0301614763.70323786.670.5596.160179263622861.040.728937055106.96320.114283752.97507456728087641200848170860.64126171.529087.12123.8322935.4751.556897351.052074.308074235116.553050.163.12043375.3231.6642.7870.2131755.3333802.615.134219.7689301988235723848.540879540819.2019.4728.7319.0185910.9515884966672170.72.62544415.96117588.99242.8519.910216.174213273.69463.6201590.50147.270358.4230219.532319.0437.54517.5218.21797.6716.050216.086918.131243407.58940612.54238180000064.959737.005113.3683853000075.469025.2423176.333687.6712.27195.7051.0811.3715.172.26114236.521.26407728.4692848933333330.10411.8654.3764208.49217.8804384040.5245.9474350.337.131047.9616027.62941.43270.380191.11131471616.32512.63207728.278135075.07212.48214.29269969428.38721.1702160.56351.29986.04247.29793.781333.7323.128.30040.23800186667202.59383.576.307.622394337.0042351900000375.73117.5385744542104.2390.0243.42139.562670033333470.02465.062530.533.75446.40119.7316906333149.378121.1926.45882455664153.12467150841.686713033362.750220.56916.075314.4463239666758.3791167.765.27146.0971126.429.056.49014.10131.781874763213876666716402666673225.728.51643.730702.614.4365376671131.120.1756.836461.6176.9154.91641.1256.3240.5527764723.3347.34711.10147.323.10129.7680190882.970944.5758135.066533.334129.99731.452801.378313.133663.0289724.553521.21795.227815.6770061.254747.24175.109144.58912380.902139.0967.2418761.825654.959921286.6144.6987.465366.695.9309571.452235.9713020091.1743966337710.60365879036900.542939.2135939.25201.109568863575341313.2112510975.832304222993401707134562133866125.1533.92142.161.774499084949277.293491173203967204472.53568.8141994357040553317313102.2856.715628751395446.4214.41617.19288129.110.8402993136.60183244.711.0273.12197603545531.032.81411664555.26169.441970141.68101189899332312775833827571691.79246822.5217252.2765.0265576.8642.353892641.653998.024521165396.223816.706.21263347.9316.5523.17127.8453196.1237127.2810.205139.3457321049123752102.6681385925416.3016.6829.9815.84172614.1031562616673787.02.17463795.6383665.75824.9935.932413.08697157.961108.3972709.93040.5671107.6299199.079166.48513.39226.0014.48317.7729.501829.349626.691776189.68611613.344670966667104.842572.329719.751444066667125.749454.4054141.954492.8853.49173.9091.7217.1522.253.51199256.630.945075604.7554825133333174.67716.9142.8735565.23205.8853373323.9228.9931190.287.321604.1126947.51232.25175.625135.97649172913.73661.98293246.705174862.56134.41916.96358952246.93616.6394114.29263.331479.53441.65183.110341.5124.006.33451.911091460000373.50467.916.8112.617509246.6803254033333447.50145.2980770465111.7194.8432.14344.753596400000501.07499.18178527.51491.89141.9516904667123.882102.1220.78877501248134.78213150932.596706400054.823184.29814.402328.4173256000051.2458987.625.55122.0891040.227.265.99111.89732.481993763249405000019632333333207.024.72743.776651.913.8364976671041.818.4747.121435.2169.6475.31738.6536.0230.5427758922.3445.29710.91645.953.09129.71379772034.052705.0226740.453436.228327.60742.026591.652706.005255.7449130.473324.65107.348487.0411172.801860.11205.749214.85818460.673121.6708.3996276.422769.450130236.0122.61710.1710970.452.8810991.886797.3006519876.5962441720910.5376781220561.423250.269776.41410.290055184043676319.882466122.3188939807832433477915279144212.9958.15959.691.5426374069006712.79975821877961744402.67301.302440706742344699397207.7462.53334076813667.6384.38419.44162727.418.1021855135.32609876.870.6225.332160742425441.542.27113379097.80290.563923851.56592014115453747710702660894.87131606.869686.39107.2372877.3251.356254051.862243.847336635234.063076.663.62763406.9628.4938.4878.1051952.6064093.735.970222.5399901792212325831.838998817619.5819.2938.5019.2486407.4615979933332514.02.61721467.08134048.96439.1722.362516.389812410.2970.8431771.06542.274765.3318215.127301.1368.28218.2416.91668.1318.553318.485118.841356351.64416513.59276906666771.825342.983413.8994455000083.492130.5872168.859678.4832.39190.0631.1412.7215.682.38130257.501.19643705.6063222533333296.79912.5048.4874386.48213.9854315467.1741.7461920.377.761108.9518130.84342.22252.648188.93034236315.73942.28776632.157140144.01197.84413.84572225384.06320.0103146.03357.571058.99284.33363.704335.5824.947.95941.34882200000217.11398.406.827.748432346.2132739633333386.00125.1486354518103.5689.5340.50140.822972400000476.91474.682245.231.92461.62123.6916885333143.808117.8124.84083475645150.66848158852.366712766765.359215.49615.376309.0368900333356.4016987.725.77151.7181127.531.406.57214.36231.998248762230896666716797000003229.128.47983.550705.514.5365390001133.319.6506.605464.7273.0604.86739.8766.5010.5527726023.7748.14712.01348.353.10129.7735922302.887344.6475735.111033.987729.42501.472491.416623.320303.3703423.656523.66255.290275.9196163.535349.97405.260754.67269382.109155.7886.5496561.010759.179515081.0149.9968.006445.095.1141611.502145.9250118621.2736650040032.81407534240368.645195.1141954.15186.3253910338286011312.1069012115.037564522053469131156658155959117.9130.19664.911.545229124033476.663875103727352152972.79569.6049790749047375594614033.9462.116578891627941.4201.38220.24316359.29.2853279262.53034940.381.0982.69192072548534.829.19551535751.29156.701811137.91118114895735314846696270501801.89257384.5518561.9158.2485525.5638.303617937.844351.594197166492.443819.837.20754382.9715.2921.45133.9903349.7567689.3211.887544.668302961114353269.4691535440014.9214.7441.5714.96171436.4732051610004386.62.10480889.1865663.82323.4539.338912.59746840.057114.9022872.56233.9621116.6053185.46153.40714.12728.9614.45837.2633.112533.127529.621852187.22400414.425367033333112.723183.339522.141518933333129.748262.6357132.712478.4673.66169.8501.7418.7224.653.65227278.520.864358581.7695036000000161.88616.6443.9115933.65199.8963392842.3827.5772910.307.971677.6230152.24432.44171.174136.68952551413.51901.71129651.816176900.32128.06616.65560046231.03317.9873110.56269.201528.41483.10492.961308.9525.726.26652.231206066667401.45477.337.4112.578558252.6713337066667449.91138.3980139566107.3095.3930.53046.253896500000471.79467.961873.127.63460.66135.2816884000122.704106.2720.00878057728159.31171164940.406626600055.979181.62914.094325.9775717666749.1361757.466.02127.5461039.227.906.05211.96728.840157763249134000020904666673222.824.68183.738663.813.3365213331046.018.1387.038439.8468.4375.31138.1366.0770.5427729823.3046.79712.08546.593.09129.71494090823.958195.3912340.517836.003927.77722.087801.718756.517446.1779634.250729.44487.308227.5325973.968455.90195.886285.00222475.254121.7808.2601679.377069.201411866.7124.26310.4412793.012.5273631.767782.5631921583.21122332915943.1783389217891.621314.271245.26420.64676180699452316.030805808.8222199277232878629451394551207.2051.820100.395.8531612642117312.92934132250388507303.58298.562761796172653661705943.6655.99401278974189.9373.18217.84179647.515.2112194630.60044968.610.5724.576174814627431.146.84913138087.92261.973664954.847002393527177688046936971068.43135125.159004.2393.3483325.9056.015802656.441643.436800633811.632865.724.26898088.1125.8934.8386.9672174.1923606.127.088523.5437051657197723072.3021069536020.2520.4662.4620.1383448.7016050970002616.93.03922477.77147469.32835.9624.644518.141311980.80682.8482071.18952.816770.0540357.355313.1008.76220.2319.859320.9521.932421.893220.971734357.00971216.50324380000079.380551.297115.71102083333391.992531.0738222.029784.3042.71205.5251.2913.9517.852.71157356.381.07743852.3263500833333315.41013.4647.7764209.10227.4234481667.9542.2194190.439.791278.4521628.71639.48254.003191.56238819016.09562.14485734.892115106.30187.01714.04071227362.45420.1453140.43328.471205.30333.99023.267380.1436.207.04750.03918566667167.83461.237.107.532502326.1433200600000448.46148.4590871770120.9385.1639.45049.423181266667481.66486.052070.331.49465.50143.6716882333140.233109.3024.05687797206160.72685158969.626707466770.908210.93414.724294.3374441000053.4902767.946.43164.2501032.037.867.06115.06031.384585765253920000018905666673209.632.34333.169680.214.6365060001069.119.2266.278455.5769.9314.92337.7526.7190.5227720524.9749.84711.37849.793.08129.7968874793.065002.7969430.916529.408034.00211.273681.172923.192983.1700221.343618.93195.215204.4997662.127649.63764.866044.39649329.036116.3158.5971955.120450.352224688.5162.4748.727562.804.9558651.052742.9702135230.1784173847654.95413696934614.141773.0147311.14196.4964411910680210711.2843811263.242846254477557054183583183603115.8927.10745.783.826244029729476.391698174422025329772.50654.8455128695852504999815519.9265.027932531916162.9203.40721.73350187.57.8813902076.24361235.911.2932.32577814154601.129.14871413846.38141.541683737.44136953432847015330263007271843.61261127.0920937.5850.8315483.8137.083341137.554744.053881358128.243960.288.37017630.0613.7419.13140.8283520.7057862.5813.751845.965955894106343724.0621729702714.0814.5937.5814.03163333.2131244026674853.62.83498793.6453914.93321.1542.818812.75466674.12133.6283340.69436.2493124.7084288.471156.46115.45031.2415.555510.4438.223538.226932.042716196.80811615.046116566667132.280686.984524.711616966667152.747260.8515180.032566.3774.48188.8902.1815.9627.654.41265283.110.736170696.9805361266667177.76318.7041.5637356.63212.9243245908.4028.5560370.297.721899.6434242.28126.58172.990144.64059272914.52801.46921050.521189893.31122.31414.85667315218.57016.5705107.36214.801730.25484.54992.364331.7625.125.00867.641126466667423.18576.088.828.036617246.3353558600000528.93174.9583651785115.4792.2329.98154.883723666667501.63507.561960.928.22485.63161.0616843333118.572101.62919.50680169620162.02669182996.476699133365.224178.28213.262317.3679949666748.564188.165.72156.516963.637.046.69812.53635.604097767265550000020250000003210.130.03683.044675.515.236461000992.118.2936.172456.1867.3755.07437.3116.4800.5127547224.5648.77712.69748.763.07129.71877617694.058813.4657133.289929.753633.60821.763531.435736.390315.6251734.305427.59446.906195.5510168.829960.35895.292874.72530352.73892.347310.8362378.398964.285830108.7154.50211.1214946.762.5298071.109002.2194846877.51232130026943.4291690431177.237265.9109416.42302.06416657395792016.969108378.323886054770338146105831105781157.5648.40639.624.0834415091693710.19670662236257380132.36403.203151783812975179469515.8655.074281161036735.6290.44813.73195648.814.2572369998.35694063.610.5894.250169904631594.460.40942312580.29236.733199363.607443876836708895675065401159.66232012.7712317.9489.0704656.9263.065079863.302924.665952263760.454059.704.63035062.0723.7231.91108.1242703.0885725.037.624029.1410781441170938433.9731216841323.3323.4126.8523.49107321.8918147600003088.73.30836592.20153584.20332.0827.101020.879510602.408106.3952659.88366.516282.9663460.775308.80110.47724.2322.968916.6130.840430.679924.802004323.24569711.17316720000097.186154.454716.061178766667112.661135.1361270.329830.7253.32234.5401.589.4016.643.31167224.640.907161908.1873806833333314.05714.7443.0486757.26259.1864752673.7933.0653930.496.721486.1222834.30139.38214.431248.68244163619.67731.84753737.46081279.79165.04813.26475396316.90225.1591128.74410.591277.46327.78293.391319.7621.876.70138.021152933333284.64462.788.066.458529462.8553148233333423.59140.539461680791.5181.7334.59135.103503200000458.23455.791878.327.90437.20129.0218774333122.718105.9821.35792709751148.40248171819.897314066779.437187.50712.595280.2678377000048.4250519.805.19177.8611221.838.937.87615.96631.704905691247550000019277666673600.533.37972.971742.918.0402440001242.617.2585.784520.2264.4754.45634.7917.3920.6030989626.9353.99799.71753.923.47129.51097663452.159802.1696629.956629.018534.45951.202981.100423.237803.1839216.553415.03764.016873.6964950.830739.74514.263523.85774302.266109.6709.1188547.895343.535526378.1174.8329.388089.144.1827071.437874.6481431830.2892689348122.16518667759139.272004.0228184.68137.038021319760501209.96452616324.74652076070757341020549920677886.9325.02143.413.056842843827605.391259674440991442702.48736.2358614787157839280018375.6357.137970932057021.6155.74114.67380867.57.4024966250.77798333.611.1492.18788188161499.631.45891212441.85125.871435841.45149752993937717693131460132237.69455312.0924676.8648.0918719.3740.772424940.835570.633339198764.937437.818.90916038.9212.7617.29191.8184795.46210934.1714.633458.25334075989870826.7341758573716.0016.1728.1716.25220893.7836096463335493.43.050851128.7864181.29318.3148.757614.19805601.101199.7344993.34338.4240153.1622356.764153.92218.54040.9216.96727.4954.119553.934841.583222171.37577811.696077300000173.220897.938327.112027000000199.608969.2976236.648605.5895.09214.8122.4013.1530.375.03284234.240.624250754.8056284300000171.23119.9138.0499360.89251.3053780240.5623.5589320.466.952039.6337455.49631.90145.413195.45367423915.43411.43236256.400128566.81106.81530.20333110189.73922.4251100.00378.051681.44550.01682.659291.3722.965.11065.381493833333545.21577.768.886.856689354.7543520766667501.39169.759277163994.6682.7325.87253.664450525000438.60443.581809.726.88423.30150.3318200533103.37697.91117.38989231850139.61713194858.847049900070.426157.36311.493286.7888370125043.6953999.675.18165.4951165.839.317.36013.25629.325125692262006666721546333333586.331.71092.805755.517.8389506001189.716.1835.613519.8960.5264.57933.6506.9600.5930843426.8053.29799.49153.223.46129.52226103782.817392.8026331.532428.848134.66281.649191.324546.496935.8395431.790726.08665.109484.2953164.789144.59964.654313.98013327.875100.7389.9268470.437758.939430547.6162.97912.7416004.072.1594592.254225.8526742114.612425519OpenBenchmarking.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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Weld Porosity Detection FP16-INT8 - Device: CPUEPYC 8534PNEPYC 8534PEPYC 9654EPYC 9684XEPYC 9754Xeon Platinum 8490HEPYC 9654 2PXeon Max 9468EPYC 9684X 2PXeon Max 9480EPYC 9754 2PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2P10K20K30K40K50KSE +/- 1.69, N = 3SE +/- 0.41, N = 3SE +/- 4.46, N = 3SE +/- 11.71, N = 3SE +/- 5.50, N = 3SE +/- 2.71, N = 3SE +/- 147.69, N = 3SE +/- 234.43, N = 4SE +/- 110.20, N = 3SE +/- 21.27, N = 3SE +/- 11.27, N = 3SE +/- 47.41, N = 3SE +/- 80.58, N = 3SE +/- 233.95, N = 3SE +/- 53.98, N = 3SE +/- 39.08, N = 34830.015366.799356.719997.8911508.1915943.1718666.1619183.8320550.9320910.5322499.0326943.4237710.6040032.8147654.9548122.161. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

BRL-CAD

BRL-CAD is a cross-platform, open-source solid modeling system with built-in benchmark mode. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgVGR Performance Metric, More Is BetterBRL-CAD 7.38.2VGR Performance MetricXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Platinum 8592+EPYC 9684XEPYC 9654EPYC 9754Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P1.3M2.6M3.9M5.2M6.5M71393776781277135283389285199791690431225743161216348447736587904075342413696951866775861750616222861939211. (CXX) g++ options: -std=c++17 -pipe -fvisibility=hidden -fno-strict-aliasing -fno-common -fexceptions -ftemplate-depth-128 -m64 -ggdb3 -O3 -fipa-pta -fstrength-reduce -finline-functions -flto -ltcl8.6 -lnetpbm -lregex_brl -lz_brl -lassimp -ldl -lm -ltk8.6

ASKAP

ASKAP is a set of benchmarks from the Australian SKA Pathfinder. The principal ASKAP benchmarks are the Hogbom Clean Benchmark (tHogbomClean) and Convolutional Resamping Benchmark (tConvolve) as well as some previous ASKAP benchmarks being included as well for OpenCL and CUDA execution of tConvolve. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMpix/sec, More Is BetterASKAP 1.0Test: tConvolve MPI - DegriddingEPYC 8534PNEPYC 8534PXeon Platinum 8490HXeon Max 9468Xeon Max 9480EPYC 9754Xeon Platinum 8592+Xeon Platinum 8490H 2PXeon Max 9468 2PXeon Max 9480 2PEPYC 9654EPYC 9754 2PXeon Platinum 8592+ 2PEPYC 9684XEPYC 9654 2PEPYC 9684X 2P30K60K90K120K150KSE +/- 69.57, N = 3SE +/- 44.43, N = 3SE +/- 93.91, N = 3SE +/- 0.00, N = 3SE +/- 76.47, N = 3SE +/- 598.81, N = 12SE +/- 153.57, N = 3SE +/- 378.22, N = 4SE +/- 143.60, N = 3SE +/- 0.00, N = 3SE +/- 189.10, N = 3SE +/- 1077.92, N = 14SE +/- 480.95, N = 3SE +/- 0.00, N = 3SE +/- 882.91, N = 4SE +/- 0.00, N = 315903.416704.417891.618522.020561.428476.631177.234614.136900.540368.642172.452639.559139.264260.281025.8123480.01. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

OpenBenchmarking.orgMpix/sec, More Is BetterASKAP 1.0Test: tConvolve MPI - GriddingEPYC 8534PNEPYC 8534PXeon Platinum 8490HXeon Max 9468Xeon Max 9480EPYC 9754Xeon Platinum 8592+Xeon Platinum 8490H 2PXeon Max 9468 2PXeon Max 9480 2PEPYC 9654EPYC 9754 2PXeon Platinum 8592+ 2PEPYC 9684XEPYC 9654 2PEPYC 9684X 2P30K60K90K120K150KSE +/- 0.00, N = 3SE +/- 86.77, N = 3SE +/- 76.67, N = 3SE +/- 99.83, N = 3SE +/- 0.00, N = 3SE +/- 685.49, N = 12SE +/- 219.20, N = 3SE +/- 412.68, N = 4SE +/- 194.30, N = 3SE +/- 802.88, N = 3SE +/- 1853.72, N = 3SE +/- 1180.69, N = 14SE +/- 1080.03, N = 3SE +/- 581.17, N = 3SE +/- 2386.93, N = 4SE +/- 19579.11, N = 316399.716573.221314.221766.523250.228854.337265.941773.042939.245195.146233.754066.072004.074389.086757.7121976.91. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.CEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654EPYC 9754Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9754 2PEPYC 9684XXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2P80K160K240K320K400KSE +/- 14.31, N = 3SE +/- 23.20, N = 3SE +/- 67.02, N = 3SE +/- 9.58, N = 3SE +/- 150.79, N = 3SE +/- 31.76, N = 4SE +/- 72.03, N = 4SE +/- 126.50, N = 4SE +/- 679.14, N = 4SE +/- 420.81, N = 4SE +/- 560.20, N = 4SE +/- 2064.71, N = 6SE +/- 1178.78, N = 6SE +/- 1528.09, N = 5SE +/- 1929.89, N = 5SE +/- 1666.53, N = 652684.9352745.6458815.3669776.4171245.26109416.42125191.11127519.37135939.25141954.15147311.14214017.18225311.82228184.68256584.67390520.931. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Medium Mesh Size - Execution TimeEPYC 8534PEPYC 8534PNEPYC 9754Xeon Max 9468Xeon Platinum 8490HXeon Max 9480EPYC 9654Xeon Platinum 8592+Xeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9480 2PEPYC 9684XEPYC 9754 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2P150300450600750691.64690.74449.47440.83420.65410.29325.42302.06201.11196.50186.33178.25137.11137.04124.3093.571. (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

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: SHA256Xeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 8534PNEPYC 8534PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654Xeon Platinum 8592+ 2PEPYC 9684XEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P70000M140000M210000M280000M350000MSE +/- 27466135.54, N = 3SE +/- 33334799.97, N = 3SE +/- 159515825.53, N = 3SE +/- 23323764.84, N = 3SE +/- 64797308.82, N = 3SE +/- 10642001.84, N = 3SE +/- 126029257.64, N = 3SE +/- 70647143.49, N = 3SE +/- 66583540.43, N = 3SE +/- 38601729.93, N = 3SE +/- 29142783.86, N = 3SE +/- 146384349.91, N = 3SE +/- 34182873.86, N = 3SE +/- 210407867.05, N = 3SE +/- 160246527.97, N = 3SE +/- 80639802.67, N = 3444391851875184043676361806994523665739579206780673038774584721007886357534131033828601131191068021071264751376901319760501201384095889831612691622802516456298872762677664033202586862371. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

Pennant

Pennant is an application focused on hydrodynamics on general unstructured meshes in 2D. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: sedovbigXeon Max 9468Xeon Max 9480Xeon Platinum 8592+Xeon Platinum 8490HXeon Max 9468 2PEPYC 8534PNXeon Max 9480 2PEPYC 8534PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9684XEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P510152025SE +/- 0.001847, N = 3SE +/- 0.013353, N = 3SE +/- 0.002400, N = 3SE +/- 0.002987, N = 3SE +/- 0.003237, N = 4SE +/- 0.029749, N = 4SE +/- 0.003253, N = 4SE +/- 0.024479, N = 4SE +/- 0.016400, N = 4SE +/- 0.018471, N = 4SE +/- 0.052975, N = 7SE +/- 0.080453, N = 15SE +/- 0.074320, N = 15SE +/- 0.040626, N = 15SE +/- 0.041359, N = 15SE +/- 0.037087, N = 1521.02112019.88246016.96910016.03080013.21125012.15132012.10690011.36964011.2843809.9645265.9164585.4606605.1476583.4083373.0811082.9404561. (CXX) g++ options: -fopenmp -lmpi_cxx -lmpi

Xmrig

Xmrig is an open-source cross-platform CPU/GPU miner for RandomX, KawPow, CryptoNight and AstroBWT. This test profile is setup to measure the Xmrig CPU mining performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgH/s, More Is BetterXmrig 6.21Variant: GhostRider - Hash Count: 1MEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Platinum 8490HXeon Max 9480Xeon Platinum 8592+EPYC 9754Xeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9480 2PEPYC 9654EPYC 9684XXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P7K14K21K28K35KSE +/- 41.94, N = 3SE +/- 30.53, N = 3SE +/- 12.93, N = 3SE +/- 23.16, N = 3SE +/- 14.78, N = 3SE +/- 44.68, N = 3SE +/- 61.27, N = 3SE +/- 14.02, N = 3SE +/- 4.03, N = 3SE +/- 7.44, N = 3SE +/- 117.32, N = 3SE +/- 10.36, N = 3SE +/- 64.69, N = 3SE +/- 46.94, N = 3SE +/- 290.50, N = 3SE +/- 95.10, N = 34564.34660.05446.85808.86122.38378.39320.510975.811263.212115.014207.715786.616324.717873.126702.731157.81. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: SHA512Xeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HXeon Platinum 8592+EPYC 8534PXeon Max 9468 2PXeon Max 9480 2PEPYC 9654Xeon Platinum 8490H 2PEPYC 9684XXeon Platinum 8592+ 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P20000M40000M60000M80000M100000MSE +/- 500176.69, N = 3SE +/- 1250834.62, N = 3SE +/- 9990938.84, N = 3SE +/- 203213936.42, N = 3SE +/- 37654528.63, N = 3SE +/- 15082903.45, N = 3SE +/- 3500318.33, N = 3SE +/- 14339684.98, N = 3SE +/- 6285194.15, N = 3SE +/- 481040738.21, N = 3SE +/- 55712209.05, N = 3SE +/- 14734422.48, N = 3SE +/- 5822493.86, N = 3SE +/- 37307640.51, N = 3SE +/- 44023374.36, N = 3SE +/- 17716195.81, N = 31619926127718893980783221944709732221992772323886054770243784608203230422299337564522053396275754974284625447745865131913465207607075274663411378822257383914559712001047766330471. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

7-Zip Compression

This is a test of 7-Zip compression/decompression with its integrated benchmark feature. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Decompression RatingXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 8534PNEPYC 8534PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9684XEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P300K600K900K1200K1500KSE +/- 916.22, N = 3SE +/- 374.59, N = 3SE +/- 61.74, N = 3SE +/- 119.21, N = 3SE +/- 81.21, N = 3SE +/- 120.28, N = 3SE +/- 2598.93, N = 3SE +/- 6276.88, N = 3SE +/- 185.14, N = 3SE +/- 28320.74, N = 3SE +/- 111.53, N = 3SE +/- 1085.99, N = 3SE +/- 291.68, N = 3SE +/- 15794.26, N = 3SE +/- 38431.24, N = 3SE +/- 14188.09, N = 32077382433472878623381463383553679894017074691315570545734106426267160657808711033277111646513402261. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: BlowfishXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PEPYC 9654Xeon Platinum 8490H 2PEPYC 9684XXeon Platinum 8592+ 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P90K180K270K360K450KSE +/- 17.33, N = 3SE +/- 8.67, N = 3SE +/- 20.54, N = 3SE +/- 11.39, N = 3SE +/- 13.72, N = 3SE +/- 79.75, N = 3SE +/- 116.69, N = 3SE +/- 169.28, N = 3SE +/- 38.33, N = 3SE +/- 348.45, N = 3SE +/- 98.33, N = 3SE +/- 649.08, N = 3SE +/- 1.33, N = 3SE +/- 3455.48, N = 3SE +/- 454.86, N = 3SE +/- 2397.15, N = 3678937915293473945131021981058311345621566581665411835831925742054992201883164923745894154121. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: bcryptXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PEPYC 9654Xeon Platinum 8490H 2PEPYC 9684XXeon Platinum 8592+ 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P90K180K270K360K450KSE +/- 8.84, N = 3SE +/- 8.67, N = 3SE +/- 0.00, N = 3SE +/- 10.67, N = 3SE +/- 22.28, N = 3SE +/- 46.28, N = 3SE +/- 866.68, N = 3SE +/- 576.80, N = 3SE +/- 20.00, N = 3SE +/- 292.24, N = 3SE +/- 154.46, N = 3SE +/- 288.38, N = 3SE +/- 1.33, N = 3SE +/- 710.46, N = 3SE +/- 1315.55, N = 3SE +/- 4958.68, N = 4678757914493472945511022221057811338661559591665411836031928982067782201883205273762684127231. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Chrysler Neon 1MEPYC 8534PNEPYC 8534PEPYC 9754Xeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 9654Xeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9754 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684XEPYC 9684X 2P70140210280350SE +/- 0.50, N = 3SE +/- 0.47, N = 3SE +/- 1.11, N = 3SE +/- 0.08, N = 3SE +/- 0.18, N = 3SE +/- 0.15, N = 3SE +/- 1.51, N = 8SE +/- 0.20, N = 3SE +/- 0.13, N = 3SE +/- 0.14, N = 3SE +/- 0.11, N = 3SE +/- 1.00, N = 10SE +/- 0.14, N = 3SE +/- 0.71, N = 12SE +/- 0.49, N = 3SE +/- 0.46, N = 3300.69297.25231.96228.74212.99207.20176.14157.56125.15117.91115.8993.9986.9377.6477.3249.55

Primesieve

Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 12.1Length: 1e13Xeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HXeon Platinum 8592+EPYC 8534PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9684XEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P1530456075SE +/- 0.07, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.07, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 4SE +/- 0.02, N = 4SE +/- 0.01, N = 467.3258.1654.3951.8248.4145.9533.9230.2027.1125.0224.2423.2721.7612.6312.1311.561. (CXX) g++ options: -O3

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Detection FP16 - Device: CPUEPYC 9754EPYC 8534PEPYC 9654EPYC 8534PNEPYC 9684XXeon Platinum 8490HEPYC 9754 2PXeon Max 9480 2PEPYC 9654 2PXeon Max 9480Xeon Platinum 8490H 2PEPYC 9684X 2PXeon Platinum 8592+ 2PXeon Max 9468 2PXeon Max 9468Xeon Platinum 8592+50100150200250SE +/- 6.97, N = 15SE +/- 0.19, N = 3SE +/- 2.90, N = 15SE +/- 0.41, N = 3SE +/- 1.22, N = 15SE +/- 0.01, N = 3SE +/- 1.90, N = 15SE +/- 0.06, N = 3SE +/- 0.44, N = 15SE +/- 0.10, N = 3SE +/- 0.05, N = 3SE +/- 0.09, N = 3SE +/- 0.09, N = 3SE +/- 0.10, N = 3SE +/- 0.18, N = 3SE +/- 0.09, N = 3230.24161.48157.45156.59110.91100.3985.8664.9163.9859.6945.7843.7743.4142.1641.9339.62MIN: 65.18 / MAX: 718.24MIN: 55.89 / MAX: 223.7MIN: 52.3 / MAX: 282.08MIN: 61.92 / MAX: 224.56MIN: 43.17 / MAX: 198.04MIN: 41.76 / MAX: 139.86MIN: 47.19 / MAX: 207.44MIN: 49.88 / MAX: 107.03MIN: 39.78 / MAX: 114.18MIN: 42.86 / MAX: 84.25MIN: 33.94 / MAX: 85.21MIN: 37.7 / MAX: 90.56MIN: 27.3 / MAX: 82.85MIN: 32.47 / MAX: 70.84MIN: 27.14 / MAX: 54.05MIN: 26.71 / MAX: 62.241. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Llamafile

Mozilla's Llamafile allows distributing and running large language models (LLMs) as a single file. Llamafile aims to make open-source LLMs more accessible to developers and users. Llamafile supports a variety of models, CPUs and GPUs, and other options. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.7Test: wizardcoder-python-34b-v1.0.Q6_K - Acceleration: CPUXeon Max 9468Xeon Max 9480Xeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8592+ 2PXeon Platinum 8490H 2PXeon Platinum 8592+EPYC 9754 2PEPYC 8534PNEPYC 8534PEPYC 9654 2PEPYC 9684X 2PXeon Platinum 8490HEPYC 9754EPYC 9654EPYC 9684X246810SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.12, N = 6SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 31.471.541.541.773.053.824.084.635.115.165.245.245.857.478.358.48

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: ChaCha20Xeon Max 9468Xeon Max 9480EPYC 8534PNEPYC 8534PXeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PEPYC 9654Xeon Max 9480 2PEPYC 9684XXeon Platinum 8490H 2PEPYC 9754Xeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P300000M600000M900000M1200000M1500000MSE +/- 415487944.57, N = 3SE +/- 312824840.64, N = 3SE +/- 71110585.03, N = 3SE +/- 9570566.28, N = 3SE +/- 232288636.76, N = 3SE +/- 3335467.26, N = 3SE +/- 329510485.23, N = 3SE +/- 151272595.84, N = 3SE +/- 863978337.15, N = 3SE +/- 94935214.36, N = 3SE +/- 271318554.29, N = 3SE +/- 170968063.89, N = 3SE +/- 303075210.03, N = 3SE +/- 80233613.02, N = 3SE +/- 874940734.18, N = 3SE +/- 63368290.15, N = 32265973541532637406900672820533270773070689383773161264211733441509169374499084949275033110211235229124033475449040847736244029729476563963293876842843827601001502634860108661819195013069285075731. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

Xcompact3d Incompact3d

Xcompact3d Incompact3d is a Fortran-MPI based, finite difference high-performance code for solving the incompressible Navier-Stokes equation and as many as you need scalar transport equations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 193 Cells Per DirectionEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Platinum 8490HXeon Max 9480Xeon Platinum 8592+EPYC 9754EPYC 9654Xeon Max 9468 2PEPYC 9684XXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P510152025SE +/- 0.10237421, N = 3SE +/- 0.01040414, N = 3SE +/- 0.04445674, N = 4SE +/- 0.01004100, N = 4SE +/- 0.01708632, N = 4SE +/- 0.02923582, N = 4SE +/- 0.03951909, N = 5SE +/- 0.06577912, N = 5SE +/- 0.01741826, N = 5SE +/- 0.05847189, N = 6SE +/- 0.05695382, N = 15SE +/- 0.04312801, N = 5SE +/- 0.03329814, N = 5SE +/- 0.01633300, N = 4SE +/- 0.04312156, N = 5SE +/- 0.00891960, N = 521.1780307020.9379139013.9692376012.9293413012.7997582010.196706608.872259148.253835117.293491176.881212796.663875106.391698175.391259674.510618934.260785963.753199011. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: ChaCha20-Poly1305Xeon Max 9468Xeon Max 9480EPYC 8534PNEPYC 8534PXeon Platinum 8592+Xeon Platinum 8490HXeon Max 9468 2PEPYC 9654Xeon Max 9480 2PEPYC 9684XXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P200000M400000M600000M800000M1000000MSE +/- 6370607.14, N = 3SE +/- 37795783.50, N = 3SE +/- 65233629.47, N = 3SE +/- 24234601.89, N = 3SE +/- 120899798.30, N = 3SE +/- 766713911.50, N = 3SE +/- 4148679.28, N = 3SE +/- 29295317.86, N = 3SE +/- 36941390.18, N = 3SE +/- 4478378.70, N = 3SE +/- 132227134.30, N = 3SE +/- 82161991.11, N = 3SE +/- 151303743.70, N = 3SE +/- 6119305.11, N = 3SE +/- 558033954.71, N = 3SE +/- 19082587.00, N = 31613210582131877961744401933050936432111856617272236257380132250388507303203967204473509609983573727352152973839256520774422025329774440991442704570397488706989258222237676527195479060286089901. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Weld Porosity Detection FP16-INT8 - Device: CPUEPYC 8534PNEPYC 8534PEPYC 9754 2PEPYC 9754EPYC 9654 2PEPYC 9654EPYC 9684XEPYC 9684X 2PXeon Platinum 8490HXeon Max 9480 2PXeon Max 9480Xeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9468Xeon Platinum 8592+ 2PXeon Platinum 8592+3691215SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.08, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 4SE +/- 0.00, N = 3SE +/- 0.00, N = 313.1111.7811.2010.9210.1710.099.429.253.582.792.672.532.502.502.482.36MIN: 6.01 / MAX: 25.62MIN: 6.2 / MAX: 24.38MIN: 9.95 / MAX: 64.26MIN: 4.98 / MAX: 37.58MIN: 8.74 / MAX: 41.95MIN: 4.29 / MAX: 26.52MIN: 4.22 / MAX: 26.07MIN: 8.35 / MAX: 44.99MIN: 1.4 / MAX: 12.71MIN: 2.44 / MAX: 34.21MIN: 2.39 / MAX: 34.22MIN: 2.23 / MAX: 31.44MIN: 2.28 / MAX: 40.8MIN: 2.22 / MAX: 27.89MIN: 1.98 / MAX: 34.01MIN: 2.02 / MAX: 21.331. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Detection FP16 - Device: CPUEPYC 8534PEPYC 8534PNEPYC 9754Xeon Max 9468Xeon Platinum 8490HXeon Max 9480EPYC 9654Xeon Platinum 8592+EPYC 9684XXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9754 2PEPYC 9684X 2P2004006008001000SE +/- 0.25, N = 3SE +/- 0.56, N = 3SE +/- 13.40, N = 15SE +/- 1.23, N = 3SE +/- 0.04, N = 3SE +/- 0.52, N = 3SE +/- 7.13, N = 15SE +/- 0.95, N = 3SE +/- 5.20, N = 15SE +/- 1.28, N = 3SE +/- 0.55, N = 3SE +/- 0.69, N = 3SE +/- 1.42, N = 3SE +/- 5.47, N = 15SE +/- 20.94, N = 15SE +/- 2.17, N = 3197.91204.05283.18285.95298.56301.30306.27403.20433.06568.81569.60654.84736.23750.11750.961095.791. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random ReadXeon Max 9468EPYC 8534PNXeon Max 9480EPYC 8534PXeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PEPYC 9654Xeon Max 9480 2PEPYC 9684XXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P300M600M900M1200M1500MSE +/- 3698537.65, N = 15SE +/- 70537.88, N = 3SE +/- 2763560.64, N = 13SE +/- 109015.29, N = 3SE +/- 168680.90, N = 3SE +/- 242146.75, N = 3SE +/- 8491098.40, N = 15SE +/- 66268.38, N = 3SE +/- 2955608.21, N = 3SE +/- 205170.43, N = 3SE +/- 5393159.45, N = 3SE +/- 5399265.88, N = 15SE +/- 1606748.01, N = 3SE +/- 955915.20, N = 3SE +/- 5954964.38, N = 3SE +/- 2399164.12, N = 3211457048242940110244070674274091778276179617315178381419943570428557054497907490537633034551286958586147871589951070857687179105954653311699473851. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Random ReadXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PEPYC 9654Xeon Max 9480 2PEPYC 9684XXeon Platinum 8490H 2PEPYC 9754Xeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P200M400M600M800M1000MSE +/- 81327.78, N = 3SE +/- 86315.43, N = 3SE +/- 211259.57, N = 3SE +/- 3338569.32, N = 3SE +/- 206014.27, N = 3SE +/- 1119855.47, N = 3SE +/- 1120857.00, N = 3SE +/- 610621.13, N = 3SE +/- 785893.13, N = 3SE +/- 326038.01, N = 3SE +/- 1411218.60, N = 3SE +/- 153963.59, N = 3SE +/- 82996.27, N = 3SE +/- 599126.55, N = 3SE +/- 2927744.32, N = 3SE +/- 1154820.73, N = 3205346154234469939234992636265366170266186537297517946405533173409272949473755946523310091525049998573235452578392800815145110103120915111352065641. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Re-Identification Retail FP16 - Device: CPUEPYC 8534PNEPYC 8534PXeon Platinum 8490HEPYC 9654Xeon Max 9468EPYC 9684XXeon Max 9480EPYC 9754Xeon Platinum 8592+EPYC 9654 2PEPYC 9684X 2PXeon Max 9468 2PXeon Max 9480 2PEPYC 9754 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2P4K8K12K16K20KSE +/- 1.89, N = 3SE +/- 3.18, N = 3SE +/- 3.97, N = 3SE +/- 4.14, N = 3SE +/- 3.72, N = 3SE +/- 4.58, N = 3SE +/- 0.46, N = 3SE +/- 4.98, N = 3SE +/- 5.57, N = 3SE +/- 7.03, N = 3SE +/- 7.99, N = 3SE +/- 3.67, N = 3SE +/- 7.22, N = 3SE +/- 10.13, N = 3SE +/- 4.14, N = 3SE +/- 7.74, N = 33355.733692.675943.666554.126717.656846.797207.747862.659515.8612024.1812665.6213102.2814033.9414601.5215519.9218375.631. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16-INT8 - Device: CPUXeon Platinum 8490H 2PXeon Max 9480Xeon Max 9480 2PXeon Platinum 8592+ 2PXeon Max 9468 2PXeon Platinum 8490HXeon Max 9468Xeon Platinum 8592+EPYC 9754EPYC 8534PNEPYC 8534PEPYC 9654EPYC 9684XEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P1530456075SE +/- 0.03, N = 3SE +/- 0.21, N = 3SE +/- 0.06, N = 3SE +/- 0.28, N = 3SE +/- 0.05, N = 3SE +/- 0.14, N = 3SE +/- 0.15, N = 3SE +/- 0.02, N = 3SE +/- 0.26, N = 15SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.11, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 365.0262.5362.1157.1356.7155.9955.7455.0738.2134.7631.9927.4621.6315.8414.3011.95MIN: 55.15 / MAX: 104.43MIN: 48.23 / MAX: 78.43MIN: 47.82 / MAX: 101.26MIN: 39.58 / MAX: 104.91MIN: 45.43 / MAX: 89.46MIN: 27.63 / MAX: 76.45MIN: 44.83 / MAX: 69.99MIN: 46.78 / MAX: 80.73MIN: 15.81 / MAX: 88.01MIN: 17.72 / MAX: 54.79MIN: 16.09 / MAX: 51.04MIN: 14.03 / MAX: 53.11MIN: 14.35 / MAX: 50.29MIN: 13.68 / MAX: 55.05MIN: 11.91 / MAX: 38.94MIN: 11.25 / MAX: 37.161. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: WPA PSKXeon Max 9468Xeon Max 9480EPYC 8534PNEPYC 8534PXeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PEPYC 9654Xeon Max 9480 2PEPYC 9684XXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P300K600K900K1200K1500KSE +/- 987.33, N = 3SE +/- 1274.00, N = 3SE +/- 79.20, N = 3SE +/- 321.00, N = 3SE +/- 1599.31, N = 3SE +/- 864.17, N = 3SE +/- 5122.51, N = 3SE +/- 921.74, N = 3SE +/- 788.50, N = 3SE +/- 2683.36, N = 3SE +/- 2519.00, N = 3SE +/- 15579.94, N = 15SE +/- 478.70, N = 3SE +/- 1333.33, N = 3SE +/- 5291.50, N = 3SE +/- 4358.90, N = 32865523340763399983862124012784281165628756434756578896713227932537970938031871253333129400015510001. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt -lbz2

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgverify/s, More Is BetterOpenSSL 3.3Algorithm: RSA4096Xeon Max 9468EPYC 8534PNXeon Max 9480EPYC 8534PXeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PEPYC 9654EPYC 9684XXeon Max 9480 2PEPYC 9754Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P800K1600K2400K3200K4000KSE +/- 165.80, N = 3SE +/- 136.07, N = 3SE +/- 200.66, N = 3SE +/- 126.84, N = 3SE +/- 92.46, N = 3SE +/- 92.62, N = 3SE +/- 355.69, N = 3SE +/- 149.48, N = 3SE +/- 1576.96, N = 3SE +/- 277.84, N = 3SE +/- 444.74, N = 3SE +/- 862.96, N = 3SE +/- 6745.42, N = 3SE +/- 5.26, N = 3SE +/- 6125.37, N = 3SE +/- 2222.55, N = 3697707.2810291.8813667.6881480.8974189.91036735.61395446.41436324.91613828.71627941.41872096.11916162.92057021.62861959.83214376.33729470.81. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

RELION

RELION - REgularised LIkelihood OptimisatioN - is a stand-alone computer program for Maximum A Posteriori refinement of (multiple) 3D reconstructions or 2D class averages in cryo-electron microscopy (cryo-EM). It is developed in the research group of Sjors Scheres at the MRC Laboratory of Molecular Biology. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRELION 4.0.1Test: Basic - Device: CPUEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 9754Xeon Platinum 8592+EPYC 9654Xeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9480 2PEPYC 9754 2PEPYC 9684XXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2P110220330440550SE +/- 6.20, N = 3SE +/- 4.63, N = 3SE +/- 2.98, N = 3SE +/- 4.65, N = 9SE +/- 7.14, N = 9SE +/- 7.69, N = 9SE +/- 2.68, N = 7SE +/- 2.95, N = 3SE +/- 3.24, N = 12SE +/- 2.65, N = 3SE +/- 2.66, N = 12SE +/- 4.04, N = 12SE +/- 2.17, N = 3SE +/- 1.41, N = 3SE +/- 2.11, N = 12SE +/- 1.27, N = 15496.24468.69412.37384.38373.18360.21290.45238.88214.42203.41201.38201.27176.42155.74145.6593.171. (CXX) g++ options: -fopenmp -std=c++11 -O3 -rdynamic -ldl -ltiff -lfftw3f -lfftw3 -lpng -ljpeg -lmpi_cxx -lmpi

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16-INT8 - Device: CPUXeon Platinum 8490H 2PXeon Max 9480 2PXeon Max 9480Xeon Platinum 8490HXeon Max 9468 2PXeon Max 9468Xeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 8534PNEPYC 9754EPYC 8534PEPYC 9654EPYC 9684XEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P510152025SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.12, N = 15SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 321.7320.2419.4417.8417.1916.3314.6713.7311.4810.9710.788.587.855.004.464.08MIN: 15.08 / MAX: 39.82MIN: 13.69 / MAX: 36.9MIN: 15.1 / MAX: 32.19MIN: 9.08 / MAX: 37.11MIN: 13.16 / MAX: 42.02MIN: 13.49 / MAX: 26.98MIN: 12.73 / MAX: 41.28MIN: 12.23 / MAX: 23.01MIN: 5.61 / MAX: 26.2MIN: 4.5 / MAX: 46.56MIN: 5.12 / MAX: 22.93MIN: 3.84 / MAX: 27.82MIN: 3.77 / MAX: 30.51MIN: 4.38 / MAX: 33.93MIN: 3.61 / MAX: 26.57MIN: 3.61 / MAX: 25.591. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

QuantLib

QuantLib is an open-source library/framework around quantitative finance for modeling, trading and risk management scenarios. QuantLib is written in C++ with Boost and its built-in benchmark used reports the QuantLib Benchmark Index benchmark score. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMFLOPS, More Is BetterQuantLib 1.32Configuration: Multi-ThreadedXeon Max 9468EPYC 8534PNXeon Max 9480EPYC 8534PXeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PEPYC 9654Xeon Max 9480 2PEPYC 9684XXeon Platinum 8490H 2PEPYC 9754Xeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P160K320K480K640K800KSE +/- 155.19, N = 3SE +/- 86.28, N = 3SE +/- 40.48, N = 3SE +/- 82.80, N = 3SE +/- 105.60, N = 3SE +/- 63.14, N = 3SE +/- 286.86, N = 3SE +/- 86.78, N = 3SE +/- 160.59, N = 3SE +/- 412.72, N = 3SE +/- 362.38, N = 3SE +/- 156.35, N = 3SE +/- 248.00, N = 3SE +/- 713.18, N = 3SE +/- 1065.76, N = 3SE +/- 804.57, N = 3143755.3153920.8162727.4172566.8179647.5195648.8288129.1294183.2316359.2319827.2350187.5374413.5380867.5593246.9644091.0759922.11. (CXX) g++ options: -O3 -march=native -fPIE -pie

m-queens

A solver for the N-queens problem with multi-threading support via the OpenMP library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterm-queens 1.2Time To SolveXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 8534PNXeon Platinum 8592+EPYC 8534PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654Xeon Platinum 8592+ 2PEPYC 9684XEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P510152025SE +/- 0.031, N = 3SE +/- 0.035, N = 3SE +/- 0.025, N = 4SE +/- 0.029, N = 4SE +/- 0.030, N = 4SE +/- 0.019, N = 4SE +/- 0.021, N = 5SE +/- 0.028, N = 5SE +/- 0.021, N = 6SE +/- 0.019, N = 6SE +/- 0.028, N = 6SE +/- 0.025, N = 6SE +/- 0.021, N = 6SE +/- 0.093, N = 15SE +/- 0.064, N = 15SE +/- 0.031, N = 821.03018.10215.21114.49314.25713.25010.8409.2857.8817.7387.4026.9506.3504.6094.1333.9811. (CXX) g++ options: -fopenmp -O2 -march=native

Coremark

This is a test of EEMBC CoreMark processor benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations/Sec, More Is BetterCoremark 1.0CoreMark Size 666 - Iterations Per SecondXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654EPYC 9684XEPYC 9754Xeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9754 2PEPYC 9684X 2P2M4M6M8M10MSE +/- 1747.99, N = 3SE +/- 5202.86, N = 3SE +/- 11678.22, N = 3SE +/- 9256.47, N = 3SE +/- 1374.66, N = 3SE +/- 4635.95, N = 3SE +/- 7219.73, N = 3SE +/- 20649.43, N = 3SE +/- 12743.41, N = 3SE +/- 10132.66, N = 3SE +/- 54326.71, N = 3SE +/- 4775.53, N = 3SE +/- 9700.21, N = 3SE +/- 2306.54, N = 3SE +/- 54771.53, N = 3SE +/- 76743.24, N = 31614763.701855135.332069432.532194630.602274636.132369998.362993136.603279262.533902076.243926265.284152864.884525288.004966250.787975907.458171241.718448083.731. (CC) gcc options: -O2 -lrt" -lrt

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Classroom - Compute: CPU-OnlyXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PEPYC 9654Xeon Platinum 8490H 2PEPYC 9684XXeon Platinum 8592+ 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P20406080100SE +/- 0.20, N = 3SE +/- 0.40, N = 3SE +/- 0.18, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.11, N = 3SE +/- 0.38, N = 3SE +/- 0.25, N = 3SE +/- 0.11, N = 3SE +/- 0.05, N = 3SE +/- 0.14, N = 3SE +/- 0.22, N = 3SE +/- 0.04, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 386.6776.8775.2668.6167.6763.6144.7140.3840.2635.9135.6633.6131.6521.1218.6716.82

PostgreSQL

This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average LatencyXeon Platinum 8490H 2PXeon Platinum 8592+ 2PXeon Max 9480 2PXeon Max 9468 2PXeon Max 9480Xeon Platinum 8592+Xeon Platinum 8490HXeon Max 9468EPYC 8534PNEPYC 8534PEPYC 9754 2PEPYC 9654 2PEPYC 9754EPYC 9684X 2PEPYC 9654EPYC 9684X0.29090.58180.87271.16361.4545SE +/- 0.030, N = 12SE +/- 0.040, N = 12SE +/- 0.038, N = 9SE +/- 0.017, N = 9SE +/- 0.007, N = 3SE +/- 0.005, N = 12SE +/- 0.007, N = 3SE +/- 0.008, N = 12SE +/- 0.001, N = 3SE +/- 0.002, N = 3SE +/- 0.001, N = 3SE +/- 0.004, N = 12SE +/- 0.001, N = 3SE +/- 0.006, N = 12SE +/- 0.003, N = 3SE +/- 0.002, N = 31.2931.1491.0981.0270.6220.5890.5720.5590.4140.3840.3290.2930.2920.2890.2780.2511. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

Primesieve

Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 12.1Length: 1e12Xeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 8534PNXeon Platinum 8592+EPYC 8534PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9684XEPYC 9754EPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P246810SE +/- 0.006, N = 7SE +/- 0.010, N = 7SE +/- 0.019, N = 8SE +/- 0.003, N = 8SE +/- 0.007, N = 8SE +/- 0.006, N = 8SE +/- 0.004, N = 9SE +/- 0.005, N = 10SE +/- 0.006, N = 10SE +/- 0.008, N = 11SE +/- 0.003, N = 11SE +/- 0.004, N = 11SE +/- 0.002, N = 11SE +/- 0.010, N = 15SE +/- 0.005, N = 12SE +/- 0.003, N = 126.1605.3324.5764.5454.2503.8013.1212.6912.3252.1872.0581.9841.9211.3591.2421.1991. (CXX) g++ options: -O3

PostgreSQL

This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 1000 - Mode: Read OnlyXeon Platinum 8490H 2PXeon Platinum 8592+ 2PXeon Max 9480 2PXeon Max 9468 2PXeon Max 9480Xeon Platinum 8592+Xeon Platinum 8490HXeon Max 9468EPYC 8534PNEPYC 8534PEPYC 9754 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9654EPYC 9684X900K1800K2700K3600K4500KSE +/- 18635.21, N = 12SE +/- 30933.90, N = 12SE +/- 36627.33, N = 9SE +/- 16726.04, N = 9SE +/- 17697.00, N = 3SE +/- 13863.97, N = 12SE +/- 20350.34, N = 3SE +/- 24978.23, N = 12SE +/- 4774.62, N = 3SE +/- 11512.67, N = 3SE +/- 13473.87, N = 3SE +/- 15503.17, N = 3SE +/- 41074.86, N = 12SE +/- 71870.42, N = 12SE +/- 39970.22, N = 3SE +/- 27594.68, N = 37781418818819207259760351607424169904617481461792636241637526065003044620342249234227313480909360326139893581. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsign/s, More Is BetterOpenSSL 3.3Algorithm: RSA4096EPYC 8534PNXeon Max 9468Xeon Max 9480EPYC 8534PXeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PXeon Max 9480 2PEPYC 9684XEPYC 9754Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P20K40K60K80K100KSE +/- 12.56, N = 3SE +/- 19.45, N = 3SE +/- 20.13, N = 3SE +/- 22.17, N = 3SE +/- 9.84, N = 3SE +/- 15.15, N = 3SE +/- 29.11, N = 3SE +/- 6.92, N = 3SE +/- 83.09, N = 3SE +/- 98.71, N = 3SE +/- 27.42, N = 3SE +/- 14.34, N = 3SE +/- 56.68, N = 3SE +/- 81.38, N = 3SE +/- 247.60, N = 3SE +/- 80.37, N = 321047.522861.025441.525894.627431.131594.435964.945531.048534.848996.853966.954601.161499.671942.098237.0107830.61. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: ArcFace ResNet-100 - Device: CPU - Executor: ParallelEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754EPYC 8534PNEPYC 8534PXeon Platinum 8490H 2PXeon Max 9480 2PEPYC 9654EPYC 9684XXeon Platinum 8592+ 2PXeon Max 9468 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+1428425670SE +/- 0.06, N = 3SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.08, N = 3SE +/- 0.24, N = 3SE +/- 0.12, N = 3SE +/- 0.09, N = 3SE +/- 0.11, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.27, N = 3SE +/- 0.12, N = 3SE +/- 0.28, N = 3SE +/- 0.21, N = 3SE +/- 0.10, N = 3SE +/- 0.27, N = 311.9415.0115.0524.6126.2527.3129.1529.2029.2729.8531.4632.8140.7342.2746.8560.411. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PEPYC 9684XXeon Max 9480 2PEPYC 9754Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P9K18K27K36K45KSE +/- 31.02, N = 3SE +/- 250.73, N = 3SE +/- 28.06, N = 3SE +/- 83.47, N = 3SE +/- 62.46, N = 3SE +/- 37.49, N = 3SE +/- 44.33, N = 3SE +/- 72.42, N = 3SE +/- 39.30, N = 3SE +/- 51.39, N = 3SE +/- 33.33, N = 3SE +/- 34.10, N = 3SE +/- 7.13, N = 3SE +/- 32.74, N = 3SE +/- 17.16, N = 3SE +/- 12.34, N = 3425133881137055337903138023125188961664516603153571528614138121241042889168413

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Pabellon Barcelona - Compute: CPU-OnlyXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PEPYC 9654Xeon Platinum 8490H 2PEPYC 9684XXeon Platinum 8592+ 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P20406080100SE +/- 0.31, N = 3SE +/- 0.44, N = 3SE +/- 0.05, N = 3SE +/- 0.15, N = 3SE +/- 0.16, N = 3SE +/- 0.09, N = 3SE +/- 0.08, N = 3SE +/- 0.20, N = 3SE +/- 0.04, N = 3SE +/- 0.26, N = 3SE +/- 0.16, N = 3SE +/- 0.06, N = 3SE +/- 0.13, N = 3SE +/- 0.10, N = 3SE +/- 0.11, N = 3SE +/- 0.13, N = 3106.9697.8094.4187.9285.1280.2955.2651.2948.8846.3844.8741.8539.8326.3523.8421.66

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Barbershop - Compute: CPU-OnlyXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654Xeon Platinum 8592+ 2PEPYC 9684XEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P70140210280350SE +/- 0.12, N = 3SE +/- 1.09, N = 3SE +/- 0.07, N = 3SE +/- 0.18, N = 3SE +/- 0.06, N = 3SE +/- 0.69, N = 3SE +/- 0.23, N = 3SE +/- 0.06, N = 3SE +/- 0.42, N = 3SE +/- 0.25, N = 3SE +/- 0.21, N = 3SE +/- 0.12, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.11, N = 3SE +/- 0.07, N = 3320.11290.56263.76261.97240.13236.73169.44156.70141.54139.25125.87123.90114.7679.0567.5265.58

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.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PEPYC 9684XEPYC 9754Xeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P10K20K30K40K50KSE +/- 197.99, N = 3SE +/- 111.72, N = 3SE +/- 36.47, N = 3SE +/- 74.33, N = 3SE +/- 61.20, N = 3SE +/- 26.00, N = 3SE +/- 5.55, N = 3SE +/- 25.12, N = 3SE +/- 24.66, N = 3SE +/- 21.05, N = 3SE +/- 10.97, N = 3SE +/- 29.96, N = 3SE +/- 8.00, N = 3SE +/- 27.79, N = 3SE +/- 32.79, N = 3SE +/- 16.90, N = 348844453774283739238366493199322305197011953518129181111683714358124271051410015

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 256 - Model: ResNet-50EPYC 9754 2PEPYC 9684X 2PEPYC 9654 2PEPYC 9754EPYC 8534PNEPYC 8534PEPYC 9654EPYC 9684XXeon Platinum 8490H 2PXeon Max 9480 2PXeon Platinum 8592+ 2PXeon Max 9468 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+1428425670SE +/- 0.28, N = 12SE +/- 0.32, N = 12SE +/- 0.37, N = 12SE +/- 0.08, N = 3SE +/- 0.13, N = 3SE +/- 0.18, N = 3SE +/- 0.21, N = 3SE +/- 0.09, N = 3SE +/- 0.49, N = 3SE +/- 0.45, N = 3SE +/- 0.29, N = 3SE +/- 0.52, N = 3SE +/- 0.62, N = 3SE +/- 0.76, N = 3SE +/- 0.52, N = 3SE +/- 0.64, N = 313.1116.9717.4127.7833.3033.6834.9034.9037.4437.9141.4541.6851.5652.9754.8463.60MIN: 8.01 / MAX: 14.98MIN: 14.73 / MAX: 19.36MIN: 11.49 / MAX: 19.87MIN: 26.81 / MAX: 28.37MIN: 32.26 / MAX: 33.94MIN: 31.27 / MAX: 34.43MIN: 33.85 / MAX: 35.99MIN: 34.15 / MAX: 35.71MIN: 26.88 / MAX: 41.26MIN: 16.15 / MAX: 40.53MIN: 38.42 / MAX: 42.93MIN: 37.38 / MAX: 44.02MIN: 30.49 / MAX: 54.7MIN: 33.28 / MAX: 55.78MIN: 46.5 / MAX: 58.5MIN: 37.79 / MAX: 65.67

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: AES-256-GCMEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654EPYC 9684XEPYC 9754Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P400000M800000M1200000M1600000M2000000MSE +/- 3442274884.66, N = 3SE +/- 148092414.72, N = 3SE +/- 119857064.56, N = 3SE +/- 63026881.96, N = 3SE +/- 2848140790.67, N = 3SE +/- 3966858436.56, N = 3SE +/- 818689705.82, N = 3SE +/- 671752057.50, N = 3SE +/- 290015345.25, N = 3SE +/- 108288170.03, N = 3SE +/- 112278705.55, N = 3SE +/- 1397764228.85, N = 3SE +/- 1320809584.03, N = 3SE +/- 2225250504.62, N = 3SE +/- 1859275059.09, N = 3SE +/- 1791964708.17, N = 341589396320746230975053050745672808759201411545370023935271774438768367076769655366785972059996399753900787310118989933231181148957353136953432847014975299393771526505465813171406176108719831030267001. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: AES-128-GCMEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654EPYC 9684XEPYC 9754Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P500000M1000000M1500000M2000000M2500000MSE +/- 409615403.40, N = 3SE +/- 530918594.60, N = 3SE +/- 43545680.66, N = 3SE +/- 56355870.99, N = 3SE +/- 28627701.54, N = 3SE +/- 1253503549.61, N = 3SE +/- 164763342.54, N = 3SE +/- 570058274.06, N = 3SE +/- 49279694.68, N = 3SE +/- 268465891.11, N = 3SE +/- 697208281.49, N = 3SE +/- 159598753.20, N = 3SE +/- 1622921637.03, N = 3SE +/- 337702856.52, N = 3SE +/- 1428966819.54, N = 3SE +/- 1841260363.24, N = 3488965390857539784931373641200848170747710702660768804693697889567506540897288436123996007109267116722114187012775833827571484669627050153302630072717693131460131785006879067199035205996723166119452831. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16-INT8 - Device: CPUXeon Max 9468Xeon Max 9480EPYC 8534PNEPYC 8534PXeon Platinum 8490HXeon Platinum 8592+EPYC 9754Xeon Max 9468 2PEPYC 9654Xeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9684XXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P9001800270036004500SE +/- 2.29, N = 3SE +/- 2.94, N = 3SE +/- 0.81, N = 3SE +/- 0.65, N = 3SE +/- 2.65, N = 3SE +/- 0.47, N = 3SE +/- 11.41, N = 15SE +/- 1.56, N = 3SE +/- 6.85, N = 3SE +/- 1.74, N = 3SE +/- 0.66, N = 3SE +/- 1.09, N = 3SE +/- 11.53, N = 3SE +/- 3.90, N = 3SE +/- 5.33, N = 3SE +/- 5.40, N = 3860.64894.87918.53998.091068.431159.661671.231691.791743.491801.891843.612212.412237.693353.734013.964032.661. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: LU.CXeon Max 9468EPYC 8534PNXeon Max 9480EPYC 8534PXeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654EPYC 9754EPYC 9684XXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P130K260K390K520K650KSE +/- 71.37, N = 3SE +/- 274.93, N = 3SE +/- 37.04, N = 4SE +/- 314.42, N = 4SE +/- 130.20, N = 4SE +/- 213.44, N = 5SE +/- 441.66, N = 5SE +/- 411.20, N = 5SE +/- 398.40, N = 5SE +/- 789.76, N = 5SE +/- 2566.89, N = 15SE +/- 1849.65, N = 6SE +/- 726.00, N = 5SE +/- 4162.49, N = 6SE +/- 1879.23, N = 6SE +/- 5800.63, N = 15126171.52130839.12131606.86133284.46135125.15232012.77246822.52257384.55261127.09269047.28279397.45362511.43455312.09472931.04551955.55588329.321. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Face Detection Retail FP16-INT8 - Device: CPUEPYC 8534PNXeon Platinum 8490HXeon Max 9468EPYC 8534PXeon Max 9480Xeon Platinum 8592+EPYC 9654EPYC 9684XXeon Max 9468 2PXeon Max 9480 2PEPYC 9754Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P8K16K24K32K40KSE +/- 7.49, N = 3SE +/- 5.45, N = 3SE +/- 18.13, N = 3SE +/- 2.45, N = 3SE +/- 7.39, N = 3SE +/- 4.30, N = 3SE +/- 2.59, N = 3SE +/- 7.70, N = 3SE +/- 12.64, N = 3SE +/- 3.89, N = 3SE +/- 8.57, N = 3SE +/- 12.64, N = 3SE +/- 17.69, N = 3SE +/- 46.87, N = 3SE +/- 15.46, N = 3SE +/- 10.72, N = 38279.639004.239087.129256.119686.3912317.9416038.9017230.3517252.2718561.9119917.5420937.5824676.8631974.8835809.5938546.661. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Helsing

Helsing is an open-source POSIX vampire number generator. This test profile measures the time it takes to generate vampire numbers between varying numbers of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterHelsing 1.0-betaDigit Range: 14 digitXeon Max 9468EPYC 8534PNXeon Max 9480EPYC 8534PXeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PEPYC 9654Xeon Max 9480 2PEPYC 9684XXeon Platinum 8490H 2PEPYC 9754Xeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P306090120150SE +/- 0.53, N = 3SE +/- 0.06, N = 3SE +/- 0.12, N = 3SE +/- 0.27, N = 3SE +/- 0.26, N = 3SE +/- 0.08, N = 3SE +/- 0.11, N = 3SE +/- 0.10, N = 3SE +/- 0.42, N = 3SE +/- 0.14, N = 3SE +/- 0.06, N = 3SE +/- 0.02, N = 3SE +/- 0.07, N = 3SE +/- 0.38, N = 5SE +/- 0.42, N = 3SE +/- 0.32, N = 3123.83120.56107.24105.6893.3589.0765.0362.1758.2556.9150.8348.8548.0933.6830.6926.781. (CC) gcc options: -O2 -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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16-INT8 - Device: CPUEPYC 8534PNXeon Max 9480Xeon Max 9468EPYC 8534PXeon Platinum 8490HXeon Platinum 8592+Xeon Platinum 8490H 2PXeon Max 9480 2PEPYC 9654Xeon Max 9468 2PEPYC 9754EPYC 9684XXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P3K6K9K12K15KSE +/- 2.89, N = 3SE +/- 5.68, N = 3SE +/- 8.09, N = 3SE +/- 2.22, N = 3SE +/- 0.81, N = 3SE +/- 0.73, N = 3SE +/- 11.51, N = 3SE +/- 8.70, N = 3SE +/- 8.39, N = 3SE +/- 4.11, N = 3SE +/- 66.33, N = 15SE +/- 3.54, N = 3SE +/- 6.25, N = 3SE +/- 4.21, N = 3SE +/- 4.04, N = 3SE +/- 1.24, N = 32763.482877.322935.472943.083325.904656.925483.815525.565528.155576.865778.516017.198719.3710708.9111735.4212750.591. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 64 - Model: ResNet-50EPYC 9754 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754EPYC 8534PNEPYC 8534PEPYC 9684XEPYC 9654Xeon Platinum 8490H 2PXeon Max 9480 2PXeon Platinum 8592+ 2PXeon Max 9468 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+1428425670SE +/- 0.23, N = 9SE +/- 0.35, N = 12SE +/- 0.37, N = 12SE +/- 0.10, N = 3SE +/- 0.18, N = 3SE +/- 0.06, N = 3SE +/- 0.10, N = 3SE +/- 0.26, N = 3SE +/- 0.50, N = 3SE +/- 0.46, N = 4SE +/- 0.12, N = 3SE +/- 0.51, N = 3SE +/- 0.62, N = 3SE +/- 0.71, N = 3SE +/- 0.24, N = 3SE +/- 0.12, N = 313.6817.0717.2827.6833.2333.4934.7135.0137.0838.3040.7742.3551.3551.5556.0163.06MIN: 10.81 / MAX: 14.92MIN: 11.04 / MAX: 19.93MIN: 13.59 / MAX: 19.8MIN: 26.61 / MAX: 28.44MIN: 31.97 / MAX: 33.93MIN: 32.03 / MAX: 34.13MIN: 33.58 / MAX: 35.7MIN: 33.84 / MAX: 36.39MIN: 26.46 / MAX: 40.15MIN: 27.11 / MAX: 40.39MIN: 16.88 / MAX: 42.06MIN: 36.76 / MAX: 44.4MIN: 41.6 / MAX: 53.81MIN: 37.49 / MAX: 54.33MIN: 51.84 / MAX: 59.12MIN: 33.2 / MAX: 65.54

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.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PEPYC 9684XEPYC 9754Xeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P20K40K60K80K100KSE +/- 346.19, N = 3SE +/- 172.59, N = 3SE +/- 134.33, N = 3SE +/- 64.91, N = 3SE +/- 102.40, N = 3SE +/- 90.37, N = 3SE +/- 42.05, N = 3SE +/- 37.55, N = 3SE +/- 193.23, N = 3SE +/- 312.83, N = 12SE +/- 13.86, N = 3SE +/- 64.24, N = 3SE +/- 16.05, N = 3SE +/- 109.43, N = 3SE +/- 14.62, N = 3SE +/- 47.34, N = 378093720516897362540580265079843290389263864537383361793341124249207501785516960

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 512 - Model: ResNet-50EPYC 9754 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754EPYC 8534PNEPYC 8534PEPYC 9654EPYC 9684XXeon Platinum 8490H 2PXeon Max 9480 2PXeon Platinum 8592+ 2PXeon Max 9468 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+1428425670SE +/- 0.22, N = 12SE +/- 0.23, N = 3SE +/- 0.29, N = 12SE +/- 0.04, N = 3SE +/- 0.11, N = 3SE +/- 0.08, N = 3SE +/- 0.10, N = 3SE +/- 0.19, N = 3SE +/- 0.21, N = 3SE +/- 0.34, N = 15SE +/- 0.30, N = 15SE +/- 0.26, N = 3SE +/- 0.72, N = 3SE +/- 0.43, N = 15SE +/- 0.12, N = 3SE +/- 0.19, N = 313.7616.2416.8227.5333.1833.7634.6534.6937.5537.8440.8341.6551.0551.8656.4463.30MIN: 9.99 / MAX: 14.97MIN: 11.41 / MAX: 17.76MIN: 11.96 / MAX: 19.26MIN: 26.63 / MAX: 28.06MIN: 32.26 / MAX: 33.81MIN: 31.39 / MAX: 34.3MIN: 33.72 / MAX: 35.48MIN: 33.34 / MAX: 35.97MIN: 24.5 / MAX: 40.69MIN: 19.63 / MAX: 40.77MIN: 15.4 / MAX: 43.41MIN: 31.6 / MAX: 43.65MIN: 40.71 / MAX: 53.89MIN: 30.55 / MAX: 55.87MIN: 48.09 / MAX: 58.6MIN: 48.94 / MAX: 65.25

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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Handwritten English Recognition FP16-INT8 - Device: CPUEPYC 8534PNEPYC 8534PXeon Platinum 8490HXeon Max 9468Xeon Max 9480EPYC 9654Xeon Platinum 8592+EPYC 9684XEPYC 9754Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9684X 2PEPYC 9754 2P14002800420056007000SE +/- 0.47, N = 3SE +/- 3.51, N = 3SE +/- 1.46, N = 3SE +/- 3.41, N = 3SE +/- 4.55, N = 3SE +/- 4.84, N = 3SE +/- 0.73, N = 3SE +/- 4.06, N = 3SE +/- 7.27, N = 3SE +/- 1.09, N = 3SE +/- 2.81, N = 3SE +/- 2.46, N = 3SE +/- 9.83, N = 3SE +/- 3.43, N = 3SE +/- 13.59, N = 3SE +/- 14.68, N = 31396.781542.611643.432074.302243.842739.012924.662986.333280.253998.024351.594744.055178.295570.635766.626400.941. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PEPYC 9684XEPYC 9754Xeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P20K40K60K80K100KSE +/- 13.58, N = 3SE +/- 29.36, N = 3SE +/- 41.33, N = 3SE +/- 65.97, N = 3SE +/- 58.68, N = 3SE +/- 162.52, N = 3SE +/- 231.95, N = 3SE +/- 155.80, N = 3SE +/- 89.57, N = 3SE +/- 50.97, N = 3SE +/- 141.04, N = 3SE +/- 125.33, N = 3SE +/- 47.16, N = 3SE +/- 756.08, N = 15SE +/- 27.74, N = 3SE +/- 28.11, N = 391643842618074273366680065952250019452114434942945419713881333391279432089320071

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: CG.CEPYC 8534PNEPYC 8534PXeon Platinum 8490HXeon Max 9468Xeon Max 9480EPYC 9754EPYC 9654Xeon Platinum 8490H 2PEPYC 9684XEPYC 9754 2PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PEPYC 9654 2PEPYC 9684X 2PXeon Platinum 8592+ 2P20K40K60K80K100KSE +/- 207.19, N = 6SE +/- 126.79, N = 6SE +/- 243.41, N = 12SE +/- 313.88, N = 7SE +/- 260.16, N = 15SE +/- 480.71, N = 15SE +/- 393.67, N = 15SE +/- 178.23, N = 8SE +/- 386.70, N = 9SE +/- 545.82, N = 5SE +/- 328.18, N = 8SE +/- 545.92, N = 15SE +/- 560.41, N = 15SE +/- 779.47, N = 7SE +/- 661.05, N = 7SE +/- 271.54, N = 621684.8022229.9433811.6335116.5535234.0643799.7853182.7758128.2462071.1063265.0163760.4565396.2266492.4484477.7288487.9198764.931. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: IS.DEPYC 8534PNEPYC 8534PXeon Platinum 8490HXeon Max 9468Xeon Max 9480Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+EPYC 9654EPYC 9754EPYC 9684XXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P3K6K9K12K15KSE +/- 14.39, N = 4SE +/- 9.85, N = 4SE +/- 6.56, N = 3SE +/- 15.55, N = 4SE +/- 18.08, N = 4SE +/- 12.99, N = 4SE +/- 4.96, N = 4SE +/- 23.92, N = 4SE +/- 8.92, N = 4SE +/- 17.53, N = 5SE +/- 35.67, N = 5SE +/- 29.95, N = 5SE +/- 155.00, N = 15SE +/- 85.04, N = 8SE +/- 76.70, N = 10SE +/- 119.22, N = 62718.942746.202865.723050.163076.663816.703819.833960.284059.705229.415702.136346.157437.819771.4110139.6912296.271. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: ExhaustiveXeon Max 9468EPYC 8534PNEPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PEPYC 9684XXeon Max 9480 2PEPYC 9754Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P48121620SE +/- 0.0003, N = 3SE +/- 0.0006, N = 3SE +/- 0.0010, N = 3SE +/- 0.0011, N = 3SE +/- 0.0024, N = 3SE +/- 0.0072, N = 3SE +/- 0.0002, N = 3SE +/- 0.0033, N = 3SE +/- 0.0086, N = 3SE +/- 0.0016, N = 3SE +/- 0.0025, N = 3SE +/- 0.0164, N = 3SE +/- 0.0157, N = 3SE +/- 0.0218, N = 3SE +/- 0.0170, N = 3SE +/- 0.1160, N = 33.12043.19933.51463.62764.26894.63035.63016.21266.30277.20757.55738.37018.909110.601811.910714.10381. (CXX) g++ options: -O3 -flto -pthread

ASKAP

ASKAP is a set of benchmarks from the Australian SKA Pathfinder. The principal ASKAP benchmarks are the Hogbom Clean Benchmark (tHogbomClean) and Convolutional Resamping Benchmark (tConvolve) as well as some previous ASKAP benchmarks being included as well for OpenCL and CUDA execution of tConvolve. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 1.0Test: tConvolve MT - GriddingXeon Max 9468 2PXeon Max 9468Xeon Max 9480Xeon Max 9480 2PXeon Platinum 8592+EPYC 8534PNEPYC 8534PXeon Platinum 8592+ 2PXeon Platinum 8490H 2PXeon Platinum 8490HEPYC 9754 2PEPYC 9654EPYC 9654 2PEPYC 9754EPYC 9684X 2PEPYC 9684X3K6K9K12K15KSE +/- 23.56, N = 3SE +/- 87.89, N = 12SE +/- 32.87, N = 12SE +/- 195.28, N = 8SE +/- 54.27, N = 5SE +/- 0.88, N = 3SE +/- 0.88, N = 3SE +/- 60.72, N = 3SE +/- 75.19, N = 5SE +/- 4.15, N = 3SE +/- 21.01, N = 3SE +/- 59.24, N = 3SE +/- 141.43, N = 5SE +/- 2.09, N = 3SE +/- 204.28, N = 3SE +/- 262.30, N = 123347.933375.323406.964382.975062.075821.465826.446038.927630.068088.1111208.7012394.6012708.5012731.8014185.1014955.601. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: BMW27 - Compute: CPU-OnlyXeon Max 9468EPYC 8534PNXeon Max 9480EPYC 8534PXeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PEPYC 9654Xeon Max 9480 2PEPYC 9684XXeon Platinum 8490H 2PEPYC 9754Xeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P714212835SE +/- 0.11, N = 3SE +/- 0.05, N = 3SE +/- 0.24, N = 3SE +/- 0.04, N = 3SE +/- 0.15, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.07, N = 4SE +/- 0.08, N = 4SE +/- 0.04, N = 4SE +/- 0.08, N = 4SE +/- 0.01, N = 4SE +/- 0.03, N = 4SE +/- 0.02, N = 5SE +/- 0.03, N = 6SE +/- 0.02, N = 631.6630.0728.4927.1225.8923.7216.5515.9615.2914.2913.7412.8612.768.617.757.17

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Fishy Cat - Compute: CPU-OnlyXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PEPYC 9654Xeon Platinum 8490H 2PEPYC 9684XXeon Platinum 8592+ 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P1020304050SE +/- 0.17, N = 3SE +/- 0.04, N = 3SE +/- 0.07, N = 3SE +/- 0.11, N = 3SE +/- 0.23, N = 3SE +/- 0.06, N = 3SE +/- 0.13, N = 3SE +/- 0.18, N = 3SE +/- 0.11, N = 3SE +/- 0.08, N = 3SE +/- 0.07, N = 3SE +/- 0.04, N = 3SE +/- 0.13, N = 3SE +/- 0.05, N = 4SE +/- 0.02, N = 5SE +/- 0.04, N = 542.7838.4838.0434.8334.3531.9123.1721.4520.1519.1318.3617.2916.5811.2510.169.71

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.

OpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2Xeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 8534PNEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654EPYC 9684XXeon Platinum 8592+ 2PEPYC 9754EPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P70140210280350SE +/- 0.31, N = 3SE +/- 0.96, N = 4SE +/- 0.31, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 1.48, N = 3SE +/- 0.93, N = 3SE +/- 0.99, N = 3SE +/- 0.93, N = 3SE +/- 1.71, N = 3SE +/- 1.93, N = 15SE +/- 0.01, N = 3SE +/- 3.85, N = 3SE +/- 0.08, N = 3SE +/- 2.28, N = 370.2178.1186.9797.98107.72108.12127.85133.99140.83171.77187.68191.82232.75288.54291.28309.221. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

OpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM2Xeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 8534PNEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654EPYC 9684XXeon Platinum 8592+ 2PEPYC 9754EPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P17003400510068008500SE +/- 7.81, N = 3SE +/- 23.93, N = 4SE +/- 7.77, N = 3SE +/- 0.10, N = 3SE +/- 0.13, N = 3SE +/- 0.70, N = 3SE +/- 36.96, N = 3SE +/- 23.14, N = 3SE +/- 24.63, N = 3SE +/- 23.15, N = 3SE +/- 42.85, N = 3SE +/- 48.20, N = 15SE +/- 0.20, N = 3SE +/- 96.25, N = 3SE +/- 2.03, N = 3SE +/- 57.10, N = 31755.331952.612174.192449.452692.872703.093196.123349.763520.714294.284692.114795.465818.747213.437281.967730.471. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Person Vehicle Bike Detection FP16 - Device: CPUEPYC 8534PNEPYC 8534PXeon Platinum 8490HXeon Max 9468Xeon Max 9480EPYC 9654EPYC 9754EPYC 9684XXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2PXeon Platinum 8592+ 2P2K4K6K8K10KSE +/- 3.87, N = 3SE +/- 2.97, N = 3SE +/- 4.52, N = 3SE +/- 7.17, N = 3SE +/- 9.37, N = 3SE +/- 25.68, N = 3SE +/- 54.09, N = 3SE +/- 18.02, N = 3SE +/- 4.97, N = 3SE +/- 4.05, N = 3SE +/- 3.66, N = 3SE +/- 8.98, N = 3SE +/- 6.56, N = 3SE +/- 10.88, N = 3SE +/- 4.91, N = 3SE +/- 3.37, N = 32506.102830.123606.123802.614093.735024.325594.815608.075725.037127.287689.327862.589203.199752.8910182.1610934.171. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: Very ThoroughXeon Max 9468EPYC 8534PNEPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PEPYC 9684XXeon Max 9480 2PEPYC 9754Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P510152025SE +/- 0.0016, N = 3SE +/- 0.0012, N = 3SE +/- 0.0007, N = 3SE +/- 0.0016, N = 3SE +/- 0.0054, N = 3SE +/- 0.0294, N = 3SE +/- 0.0006, N = 3SE +/- 0.0055, N = 3SE +/- 0.0254, N = 3SE +/- 0.0022, N = 3SE +/- 0.0006, N = 3SE +/- 0.0379, N = 3SE +/- 0.0007, N = 3SE +/- 0.1965, N = 3SE +/- 0.1284, N = 3SE +/- 0.1961, N = 155.13425.24065.74205.97027.08857.62409.184410.205110.322411.887512.318713.751814.633417.009819.405222.00231. (CXX) g++ options: -O3 -flto -pthread

ACES DGEMM

This is a multi-threaded DGEMM benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterACES DGEMM 1.0Sustained Floating-Point RateXeon Max 9468EPYC 8534PNEPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PXeon Max 9480 2PEPYC 9754EPYC 9684XXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P20406080100SE +/- 0.04, N = 4SE +/- 0.10, N = 5SE +/- 0.16, N = 5SE +/- 0.09, N = 5SE +/- 0.09, N = 5SE +/- 0.08, N = 6SE +/- 0.22, N = 7SE +/- 0.06, N = 7SE +/- 0.04, N = 7SE +/- 0.17, N = 7SE +/- 0.40, N = 7SE +/- 0.28, N = 7SE +/- 0.11, N = 8SE +/- 0.10, N = 9SE +/- 0.39, N = 9SE +/- 0.19, N = 1019.7721.0021.8222.5423.5429.1439.0939.3544.6744.9745.2545.9758.2568.8778.5084.691. (CC) gcc options: -O3 -march=native -fopenmp

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.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PEPYC 9684XXeon Max 9480 2PEPYC 9754Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P5001000150020002500SE +/- 1.20, N = 3SE +/- 0.58, N = 3SE +/- 0.33, N = 3SE +/- 0.58, N = 3SE +/- 0.58, N = 3SE +/- 0.88, N = 3SE +/- 0.58, N = 3SE +/- 1.76, N = 3SE +/- 1.15, N = 3SE +/- 0.58, N = 3SE +/- 1.20, N = 3SE +/- 0.58, N = 3SE +/- 0.88, N = 3SE +/- 0.58, N = 3SE +/- 1.20, N = 3SE +/- 0.88, N = 3224020431988179216571441118010491037961954894759661561531

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PEPYC 9684XXeon Max 9480 2PEPYC 9754Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P6001200180024003000SE +/- 3.06, N = 3SE +/- 1.00, N = 3SE +/- 1.20, N = 3SE +/- 1.20, N = 3SE +/- 0.88, N = 3SE +/- 1.45, N = 3SE +/- 1.00, N = 3SE +/- 2.65, N = 3SE +/- 1.53, N = 3SE +/- 0.33, N = 3SE +/- 0.88, N = 3SE +/- 0.88, N = 3SE +/- 0.00, N = 3SE +/- 0.58, N = 3SE +/- 1.45, N = 3SE +/- 0.67, N = 3264924212357212319771709139112371225114311301063898780660630

LULESH

LULESH is the Livermore Unstructured Lagrangian Explicit Shock Hydrodynamics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgz/s, More Is BetterLULESH 2.0.3EPYC 8534PNEPYC 8534PXeon Platinum 8490HXeon Max 9468EPYC 9754Xeon Max 9480EPYC 9654EPYC 9684XXeon Platinum 8592+Xeon Platinum 8490H 2PEPYC 9754 2PEPYC 9654 2PXeon Max 9468 2PXeon Max 9480 2PEPYC 9684X 2PXeon Platinum 8592+ 2P15K30K45K60K75KSE +/- 373.78, N = 12SE +/- 372.63, N = 12SE +/- 182.38, N = 15SE +/- 211.12, N = 6SE +/- 244.17, N = 3SE +/- 217.67, N = 6SE +/- 372.12, N = 3SE +/- 194.65, N = 15SE +/- 437.51, N = 4SE +/- 272.44, N = 4SE +/- 427.75, N = 3SE +/- 349.39, N = 15SE +/- 206.20, N = 5SE +/- 207.54, N = 5SE +/- 675.87, N = 3SE +/- 359.29, N = 316855.8916949.5923072.3023848.5424218.8325831.8426082.7127287.2638433.9743724.0646126.0848760.8252102.6753269.4754678.4970826.731. (CXX) g++ options: -O3 -fopenmp -lm -lmpi_cxx -lmpi

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Read While WritingXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 8534PNEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9684XEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P8M16M24M32M40MSE +/- 88257.34, N = 3SE +/- 93731.24, N = 7SE +/- 62510.72, N = 3SE +/- 86788.79, N = 3SE +/- 88013.69, N = 15SE +/- 143570.67, N = 4SE +/- 171807.74, N = 3SE +/- 197596.27, N = 3SE +/- 239363.92, N = 3SE +/- 195912.00, N = 15SE +/- 170909.18, N = 3SE +/- 118870.12, N = 3SE +/- 172052.60, N = 3SE +/- 746932.76, N = 15SE +/- 651901.79, N = 15SE +/- 803844.71, N = 158795408998817610695360118821031205390112168413138592541535440017297027175857371802499319595777228294263032459633307574365932041. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 256 - Model: ResNet-152EPYC 9754 2PEPYC 9684X 2PEPYC 9654 2PEPYC 9754EPYC 8534PNEPYC 8534PEPYC 9654EPYC 9684XXeon Platinum 8490H 2PXeon Max 9480 2PXeon Platinum 8592+ 2PXeon Max 9468 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+612182430SE +/- 0.12, N = 9SE +/- 0.13, N = 9SE +/- 0.13, N = 9SE +/- 0.15, N = 3SE +/- 0.04, N = 3SE +/- 0.06, N = 3SE +/- 0.11, N = 3SE +/- 0.20, N = 3SE +/- 0.13, N = 7SE +/- 0.18, N = 4SE +/- 0.16, N = 3SE +/- 0.09, N = 3SE +/- 0.15, N = 3SE +/- 0.15, N = 3SE +/- 0.08, N = 3SE +/- 0.29, N = 35.717.027.1411.0113.1813.2613.9514.0814.0814.9216.0016.3019.2019.5820.2523.33MIN: 3.18 / MAX: 6.47MIN: 3.92 / MAX: 8.02MIN: 3.67 / MAX: 7.91MIN: 10.7 / MAX: 11.43MIN: 12.94 / MAX: 13.34MIN: 12.98 / MAX: 13.46MIN: 13.45 / MAX: 14.29MIN: 13.43 / MAX: 14.62MIN: 6.98 / MAX: 15.46MIN: 8.28 / MAX: 15.85MIN: 11.43 / MAX: 16.66MIN: 11.86 / MAX: 16.95MIN: 10.93 / MAX: 20MIN: 13.13 / MAX: 20.27MIN: 17.03 / MAX: 21.21MIN: 19.78 / MAX: 24.54

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 512 - Model: ResNet-152EPYC 9754 2PEPYC 9684X 2PEPYC 9654 2PEPYC 9754EPYC 8534PEPYC 8534PNEPYC 9684XEPYC 9654Xeon Platinum 8490H 2PXeon Max 9480 2PXeon Platinum 8592+ 2PXeon Max 9468 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+612182430SE +/- 0.11, N = 9SE +/- 0.13, N = 9SE +/- 0.14, N = 9SE +/- 0.02, N = 3SE +/- 0.11, N = 3SE +/- 0.09, N = 3SE +/- 0.12, N = 3SE +/- 0.18, N = 3SE +/- 0.09, N = 3SE +/- 0.16, N = 3SE +/- 0.18, N = 3SE +/- 0.09, N = 3SE +/- 0.18, N = 7SE +/- 0.18, N = 7SE +/- 0.04, N = 3SE +/- 0.16, N = 35.757.077.2010.9113.2813.2813.9514.0514.5914.7416.1716.6819.2919.4720.4623.41MIN: 2.81 / MAX: 6.42MIN: 3.53 / MAX: 7.83MIN: 3.59 / MAX: 8.29MIN: 10.79 / MAX: 11.05MIN: 12.91 / MAX: 13.58MIN: 12.1 / MAX: 13.57MIN: 13.65 / MAX: 14.33MIN: 13.72 / MAX: 14.61MIN: 10.11 / MAX: 16.06MIN: 10.63 / MAX: 15.36MIN: 11.54 / MAX: 16.69MIN: 7.02 / MAX: 17.07MIN: 10.66 / MAX: 20.07MIN: 11.79 / MAX: 20.53MIN: 17.25 / MAX: 21.16MIN: 22.41 / MAX: 24.3

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Machine Translation EN To DE FP16 - Device: CPUEPYC 8534PNEPYC 9754EPYC 8534PEPYC 9654EPYC 9684XXeon Platinum 8490HEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2PXeon Max 9480 2PXeon Max 9480Xeon Platinum 8490H 2PXeon Max 9468 2PXeon Max 9468Xeon Platinum 8592+ 2PXeon Platinum 8592+20406080100SE +/- 0.05, N = 3SE +/- 1.05, N = 5SE +/- 0.13, N = 3SE +/- 0.77, N = 7SE +/- 0.60, N = 11SE +/- 0.03, N = 3SE +/- 0.59, N = 3SE +/- 0.44, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.34, N = 3SE +/- 0.16, N = 3SE +/- 0.07, N = 3SE +/- 0.12, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3109.00102.84100.6791.1683.1162.4656.7348.4343.1741.5738.5037.5829.9828.7328.1726.85MIN: 53.63 / MAX: 148.46MIN: 47.17 / MAX: 172.5MIN: 47.39 / MAX: 138.74MIN: 43.59 / MAX: 136.45MIN: 36.66 / MAX: 131.44MIN: 46.12 / MAX: 79.06MIN: 45.46 / MAX: 220.87MIN: 36.56 / MAX: 119.12MIN: 34.03 / MAX: 226.77MIN: 28.11 / MAX: 155.06MIN: 25.18 / MAX: 161.13MIN: 20.58 / MAX: 347.67MIN: 20.48 / MAX: 222.7MIN: 19.52 / MAX: 209.26MIN: 20.12 / MAX: 261.31MIN: 20.18 / MAX: 307.381. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 64 - Model: ResNet-152EPYC 9754 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754EPYC 8534PNEPYC 8534PEPYC 9654EPYC 9684XXeon Platinum 8490H 2PXeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8592+ 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+612182430SE +/- 0.03, N = 3SE +/- 0.13, N = 9SE +/- 0.15, N = 9SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.16, N = 3SE +/- 0.17, N = 3SE +/- 0.10, N = 3SE +/- 0.17, N = 3SE +/- 0.05, N = 3SE +/- 0.21, N = 3SE +/- 0.15, N = 12SE +/- 0.19, N = 5SE +/- 0.17, N = 12SE +/- 0.18, N = 3SE +/- 0.34, N = 35.797.047.0810.8813.1513.1913.8413.8914.0314.9615.8416.2519.0119.2420.1323.49MIN: 3.25 / MAX: 6.39MIN: 3.58 / MAX: 7.91MIN: 3.61 / MAX: 8.09MIN: 10.75 / MAX: 11.02MIN: 12.52 / MAX: 13.35MIN: 12.44 / MAX: 13.62MIN: 13.51 / MAX: 14.34MIN: 13.54 / MAX: 14.15MIN: 8.98 / MAX: 15.41MIN: 10.89 / MAX: 15.42MIN: 6.82 / MAX: 16.51MIN: 6.19 / MAX: 17.5MIN: 11.15 / MAX: 20.08MIN: 10.95 / MAX: 20.39MIN: 17.06 / MAX: 21.01MIN: 19.73 / MAX: 24.56

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: MG.CEPYC 8534PNEPYC 8534PXeon Platinum 8490HXeon Max 9468Xeon Max 9480Xeon Platinum 8592+EPYC 9654EPYC 9754EPYC 9684XXeon Platinum 8490H 2PXeon Max 9480 2PXeon Max 9468 2PEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9684X 2PEPYC 9754 2P50K100K150K200K250KSE +/- 301.54, N = 8SE +/- 167.65, N = 8SE +/- 152.92, N = 9SE +/- 436.97, N = 10SE +/- 382.59, N = 10SE +/- 406.29, N = 9SE +/- 1153.71, N = 15SE +/- 884.47, N = 9SE +/- 1372.89, N = 15SE +/- 1262.04, N = 9SE +/- 1298.67, N = 9SE +/- 1415.96, N = 9SE +/- 2707.72, N = 14SE +/- 1422.07, N = 7SE +/- 7207.31, N = 15SE +/- 1776.37, N = 1162449.5563930.6483448.7085910.9586407.46107321.89116725.04125567.07134994.14163333.21171436.47172614.10194942.32220893.78244738.81245447.721. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

Algebraic Multi-Grid Benchmark

AMG is a parallel algebraic multigrid solver for linear systems arising from problems on unstructured grids. The driver provided with AMG builds linear systems for various 3-dimensional problems. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFigure Of Merit, More Is BetterAlgebraic Multi-Grid Benchmark 1.2EPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9754EPYC 9654EPYC 9684XEPYC 9754 2PXeon Platinum 8490H 2PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2P1000M2000M3000M4000M5000MSE +/- 291866.98, N = 3SE +/- 2902028.33, N = 3SE +/- 3359086.30, N = 3SE +/- 1778769.08, N = 3SE +/- 468935.32, N = 3SE +/- 4000850.45, N = 3SE +/- 433397.56, N = 3SE +/- 1066998.75, N = 3SE +/- 4963892.66, N = 3SE +/- 235140242.43, N = 12SE +/- 2438223.49, N = 3SE +/- 7233468.75, N = 3SE +/- 9164253.67, N = 3SE +/- 9014927.20, N = 3SE +/- 8808919.44, N = 3SE +/- 6340323.98, N = 311471080001150052667158849666715979933331605097000181476000022753526672282802000237614900027830390833124402667315626166732051610003609646333432493033344672813331. (CC) gcc options: -lparcsr_ls -lparcsr_mv -lseq_mv -lIJ_mv -lkrylov -lHYPRE_utilities -lm -fopenmp -lmpi

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PUSCH Processor Benchmark, Throughput TotalXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 8534PNXeon Platinum 8592+EPYC 8534PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654Xeon Platinum 8592+ 2PEPYC 9684XEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P2K4K6K8K10KSE +/- 0.09, N = 3SE +/- 18.30, N = 3SE +/- 18.43, N = 3SE +/- 19.02, N = 13SE +/- 18.09, N = 3SE +/- 25.25, N = 3SE +/- 26.47, N = 3SE +/- 0.15, N = 3SE +/- 47.68, N = 6SE +/- 0.20, N = 3SE +/- 0.12, N = 3SE +/- 55.17, N = 3SE +/- 63.64, N = 5SE +/- 74.95, N = 4SE +/- 2.29, N = 3SE +/- 76.45, N = 32170.72514.02616.92818.53088.73193.03787.04386.64853.65239.65493.45656.95973.86835.76999.78396.7MIN: 1388.2 / MAX: 2170.8MIN: 1550.4 / MAX: 2532.5MIN: 1591.7 / MAX: 2653.8MIN: 1696.8 / MAX: 2893.2MIN: 1852.6 / MAX: 3124.9MIN: 1941.2 / MAX: 3219MIN: 2545.7 / MAX: 3839.9MIN: 2970.2 / MAX: 4386.9MIN: 3181.8 / MAX: 4929.4MIN: 3593.7 / MAX: 5240MIN: 3703.4 / MAX: 5493.6MIN: 3595.4 / MAX: 5767.2MIN: 4286.9 / MAX: 6228.4MIN: 6107.2 / MAX: 6950.8MIN: 6424.8 / MAX: 7003.9MIN: 8064.1 / MAX: 8549.41. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -mavx512f -mavx512cd -mavx512bw -mavx512dq -fno-trapping-math -fno-math-errno -ldl

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: fcn-resnet101-11 - Device: CPU - Executor: ParallelEPYC 9684X 2PEPYC 9654 2PEPYC 9654EPYC 9684XXeon Max 9480 2PXeon Max 9468 2PEPYC 8534PNEPYC 9754 2PEPYC 8534PXeon Max 9480Xeon Max 9468Xeon Platinum 8490H 2PXeon Platinum 8490HXeon Platinum 8592+ 2PEPYC 9754Xeon Platinum 8592+0.74441.48882.23322.97763.722SE +/- 0.001170, N = 3SE +/- 0.016597, N = 12SE +/- 0.013875, N = 12SE +/- 0.020352, N = 15SE +/- 0.026516, N = 3SE +/- 0.027020, N = 13SE +/- 0.018603, N = 15SE +/- 0.034519, N = 3SE +/- 0.023203, N = 3SE +/- 0.015170, N = 3SE +/- 0.011894, N = 3SE +/- 0.007683, N = 3SE +/- 0.008707, N = 3SE +/- 0.038257, N = 3SE +/- 0.018814, N = 3SE +/- 0.008053, N = 30.8576850.9170431.2513501.2517902.1048002.1746302.2545302.5170902.5727802.6172102.6254402.8349803.0392203.0508503.1315603.3083601. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Machine Translation EN To DE FP16 - Device: CPUEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 9654EPYC 9684XXeon Platinum 8592+EPYC 9754Xeon Platinum 8490H 2PXeon Max 9468 2PXeon Max 9480 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2PXeon Platinum 8592+ 2P2004006008001000SE +/- 0.13, N = 3SE +/- 0.39, N = 3SE +/- 1.21, N = 3SE +/- 4.09, N = 3SE +/- 0.17, N = 3SE +/- 4.59, N = 7SE +/- 4.29, N = 11SE +/- 0.54, N = 3SE +/- 6.34, N = 5SE +/- 3.48, N = 3SE +/- 1.99, N = 3SE +/- 1.13, N = 3SE +/- 8.96, N = 3SE +/- 0.37, N = 3SE +/- 11.78, N = 3SE +/- 0.85, N = 3293.07317.38415.96467.08477.77525.88576.98592.20621.42793.64795.63889.18990.161110.491127.051128.781. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Aircrack-ng

Aircrack-ng is a tool for assessing WiFi/WLAN network security. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgk/s, More Is BetterAircrack-ng 1.7Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PXeon Max 9480 2PXeon Max 9468 2PXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HXeon Platinum 8592+EPYC 8534PEPYC 9754EPYC 9654EPYC 9684X40K80K120K160K200KSE +/- 1197.41, N = 15SE +/- 773.70, N = 3SE +/- 1131.21, N = 15SE +/- 1027.07, N = 15SE +/- 36.42, N = 3SE +/- 122.31, N = 3SE +/- 319.60, N = 3SE +/- 1953.08, N = 3SE +/- 840.59, N = 3SE +/- 338.71, N = 3SE +/- 714.86, N = 3SE +/- 207.11, N = 3SE +/- 1622.56, N = 353914.9364181.2965663.8283665.76117588.99134048.96141831.78147469.33153584.20153818.03177989.89186675.43206619.291. (CXX) g++ options: -std=gnu++17 -O3 -fvisibility=hidden -fcommon -rdynamic -lnl-3 -lnl-genl-3 -lpcre -lsqlite3 -lpthread -lz -lssl -lcrypto -lhwloc -ldl -lm -pthread

EPYC 9654 2P: The test run did not produce a result.

EPYC 9684X 2P: The test run did not produce a result.

EPYC 9754 2P: The test run did not produce a result.

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Junkshop - Compute: CPU-OnlyXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654EPYC 9684XXeon Platinum 8592+ 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P1020304050SE +/- 0.04, N = 3SE +/- 0.10, N = 3SE +/- 0.03, N = 3SE +/- 0.24, N = 3SE +/- 0.06, N = 3SE +/- 0.10, N = 3SE +/- 0.06, N = 3SE +/- 0.19, N = 3SE +/- 0.04, N = 3SE +/- 0.06, N = 3SE +/- 0.13, N = 3SE +/- 0.09, N = 3SE +/- 0.08, N = 3SE +/- 0.02, N = 4SE +/- 0.08, N = 4SE +/- 0.02, N = 442.8539.1738.8635.9635.8232.0824.9923.4521.1521.0218.6518.3117.7012.8711.4711.21

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.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_timeEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654EPYC 9684XXeon Max 9468 2PEPYC 9754Xeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P1428425670SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.09, N = 3SE +/- 0.02, N = 3SE +/- 0.16, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.07, N = 3SE +/- 0.14, N = 3SE +/- 0.05, N = 316.9218.4719.9122.3624.6427.1030.6134.6435.9337.7439.3442.8248.7652.6760.3763.85

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: yolov4 - Device: CPU - Executor: ParallelEPYC 9754 2PEPYC 9684X 2PEPYC 9654 2PEPYC 8534PNEPYC 9754EPYC 8534PEPYC 9654EPYC 9684XXeon Max 9480 2PXeon Platinum 8490H 2PXeon Max 9468 2PXeon Platinum 8592+ 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+510152025SE +/- 0.07368, N = 3SE +/- 0.00606, N = 3SE +/- 0.02467, N = 3SE +/- 0.03807, N = 3SE +/- 0.01474, N = 3SE +/- 0.00112, N = 3SE +/- 0.01934, N = 3SE +/- 0.00491, N = 3SE +/- 0.02010, N = 3SE +/- 0.01439, N = 3SE +/- 0.12795, N = 3SE +/- 0.11392, N = 3SE +/- 0.05802, N = 3SE +/- 0.00279, N = 3SE +/- 0.01672, N = 3SE +/- 0.13074, N = 35.557676.318286.348918.472838.635948.758149.785169.8000812.5974012.7546013.0869014.1980016.1742016.3898018.1413020.879501. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

WRF

WRF, the Weather Research and Forecasting Model, is a "next-generation mesoscale numerical weather prediction system designed for both atmospheric research and operational forecasting applications. It features two dynamical cores, a data assimilation system, and a software architecture supporting parallel computation and system extensibility." Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterWRF 4.2.2Input: conus 2.5kmEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 9754Xeon Platinum 8592+EPYC 9654EPYC 9684XEPYC 9754 2PXeon Max 9468 2PEPYC 9654 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9684X 2P4K8K12K16K20K16609.7816147.0013273.6912410.2911980.8111895.4110602.4110176.867885.167365.867157.966918.026840.066674.125601.104431.341. (F9X) gfortran options: -O2 -ftree-vectorize -funroll-loops -ffree-form -fconvert=big-endian -frecord-marker=4 -fallow-invalid-boz -lesmf_time -lwrfio_nf -lnetcdff -lnetcdf -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

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.

OpenBenchmarking.orgBillion Interactions/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1Xeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 8534PNXeon Platinum 8592+EPYC 8534PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654EPYC 9684XXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9654 2PEPYC 9754EPYC 9684X 2P50100150200250SE +/- 0.12, N = 4SE +/- 0.84, N = 15SE +/- 0.61, N = 5SE +/- 0.03, N = 6SE +/- 0.20, N = 6SE +/- 0.04, N = 6SE +/- 1.28, N = 12SE +/- 1.85, N = 12SE +/- 0.81, N = 7SE +/- 1.18, N = 7SE +/- 3.27, N = 13SE +/- 1.02, N = 8SE +/- 3.98, N = 15SE +/- 2.97, N = 13SE +/- 0.70, N = 9SE +/- 6.15, N = 1263.6270.8482.8598.21106.40107.35108.40114.90133.63136.95151.95199.73214.62217.85229.91236.611. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

OpenBenchmarking.orgGFInst/s, More Is BetterminiBUDE 20210901Implementation: OpenMP - Input Deck: BM1Xeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 8534PNXeon Platinum 8592+EPYC 8534PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654EPYC 9684XXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9654 2PEPYC 9754EPYC 9684X 2P13002600390052006500SE +/- 2.91, N = 4SE +/- 21.10, N = 15SE +/- 15.23, N = 5SE +/- 0.82, N = 6SE +/- 5.07, N = 6SE +/- 0.97, N = 6SE +/- 32.01, N = 12SE +/- 46.19, N = 12SE +/- 20.15, N = 7SE +/- 29.52, N = 7SE +/- 81.83, N = 13SE +/- 25.44, N = 8SE +/- 99.42, N = 15SE +/- 74.36, N = 13SE +/- 17.54, N = 9SE +/- 153.85, N = 121590.501771.072071.192455.162659.882683.832709.932872.563340.693423.633798.704993.345365.425446.165747.825915.251. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: ArcFace ResNet-100 - Device: CPU - Executor: StandardEPYC 9754 2PEPYC 9684X 2PEPYC 9654 2PEPYC 8534PEPYC 9754EPYC 8534PNXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654Xeon Max 9468 2PEPYC 9684XXeon Max 9480Xeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+1530456075SE +/- 0.41, N = 15SE +/- 0.57, N = 15SE +/- 0.32, N = 15SE +/- 0.07, N = 3SE +/- 0.38, N = 15SE +/- 0.29, N = 15SE +/- 0.21, N = 3SE +/- 0.48, N = 3SE +/- 0.51, N = 13SE +/- 0.42, N = 4SE +/- 0.34, N = 3SE +/- 0.06, N = 3SE +/- 0.59, N = 3SE +/- 0.69, N = 15SE +/- 0.16, N = 3SE +/- 0.82, N = 318.0123.5023.5830.3730.8631.0633.9636.2538.4239.2640.5741.4742.2747.2752.8266.521. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer ISPC - Model: CrownEPYC 8534PNXeon Max 9468EPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PEPYC 9654Xeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9754EPYC 9684XXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9754 2PEPYC 9684X 2P50100150200250SE +/- 0.07, N = 4SE +/- 0.05, N = 4SE +/- 0.04, N = 4SE +/- 0.02, N = 4SE +/- 0.07, N = 5SE +/- 0.05, N = 5SE +/- 0.14, N = 6SE +/- 0.08, N = 6SE +/- 0.48, N = 6SE +/- 0.12, N = 6SE +/- 0.64, N = 6SE +/- 0.20, N = 7SE +/- 0.07, N = 7SE +/- 0.43, N = 7SE +/- 1.04, N = 7SE +/- 1.23, N = 858.2858.4262.1965.3370.0582.97107.63110.53116.61124.71126.15130.79153.16176.21190.08214.02MIN: 57.28 / MAX: 60.13MIN: 55.81 / MAX: 61.43MIN: 61.2 / MAX: 63.88MIN: 60.89 / MAX: 69.91MIN: 63.26 / MAX: 79.13MIN: 79.09 / MAX: 87.11MIN: 102.22 / MAX: 114.8MIN: 107.72 / MAX: 115.26MIN: 107.47 / MAX: 129.72MIN: 115.3 / MAX: 143.82MIN: 122.27 / MAX: 133.25MIN: 126.46 / MAX: 138.83MIN: 143.78 / MAX: 166.08MIN: 163.69 / MAX: 195.97MIN: 164.86 / MAX: 219.06MIN: 181.82 / MAX: 248.2

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: T5 Encoder - Device: CPU - Executor: StandardEPYC 8534PEPYC 8534PNEPYC 9754 2PEPYC 9754EPYC 9654 2PXeon Max 9480 2PEPYC 9654Xeon Max 9468 2PXeon Max 9480Xeon Max 9468EPYC 9684X 2PEPYC 9684XXeon Platinum 8490H 2PXeon Platinum 8592+ 2PXeon Platinum 8490HXeon Platinum 8592+100200300400500SE +/- 0.98, N = 13SE +/- 0.54, N = 3SE +/- 1.15, N = 3SE +/- 1.16, N = 3SE +/- 1.86, N = 3SE +/- 0.89, N = 3SE +/- 1.08, N = 3SE +/- 1.31, N = 3SE +/- 0.51, N = 3SE +/- 3.95, N = 15SE +/- 4.17, N = 15SE +/- 1.35, N = 3SE +/- 0.26, N = 3SE +/- 3.01, N = 3SE +/- 0.84, N = 3SE +/- 1.00, N = 3127.09127.21127.54131.00167.11185.46187.01199.08215.13219.53263.52279.64288.47356.76357.36460.781. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: ResNet50 v1-12-int8 - Device: CPU - Executor: ParallelEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754Xeon Max 9480 2PXeon Platinum 8592+ 2PXeon Platinum 8490H 2PXeon Max 9468 2PEPYC 9684XEPYC 9654EPYC 8534PNEPYC 8534PXeon Max 9480Xeon Platinum 8592+Xeon Platinum 8490HXeon Max 946870140210280350SE +/- 0.66, N = 3SE +/- 0.54, N = 3SE +/- 0.54, N = 3SE +/- 0.53, N = 3SE +/- 0.73, N = 3SE +/- 1.63, N = 3SE +/- 0.73, N = 3SE +/- 0.47, N = 3SE +/- 1.50, N = 3SE +/- 2.31, N = 4SE +/- 1.50, N = 15SE +/- 1.54, N = 15SE +/- 2.72, N = 3SE +/- 2.50, N = 3SE +/- 0.53, N = 3SE +/- 1.74, N = 388.37100.36100.46153.25153.41153.92156.46166.49178.70190.38203.06207.34301.14308.80313.10319.041. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

GROMACS

The GROMACS (GROningen MAchine for Chemical Simulations) molecular dynamics package testing with the water_GMX50 data. This test profile allows selecting between CPU and GPU-based GROMACS builds. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2024Implementation: MPI CPU - Input: water_GMX50_bareEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654EPYC 9754Xeon Max 9468 2PXeon Max 9480 2PEPYC 9684XXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9754 2PEPYC 9684X 2P510152025SE +/- 0.004, N = 3SE +/- 0.004, N = 3SE +/- 0.019, N = 3SE +/- 0.007, N = 3SE +/- 0.010, N = 3SE +/- 0.009, N = 3SE +/- 0.034, N = 3SE +/- 0.032, N = 3SE +/- 0.025, N = 3SE +/- 0.032, N = 3SE +/- 0.052, N = 3SE +/- 0.019, N = 3SE +/- 0.046, N = 3SE +/- 0.189, N = 6SE +/- 0.067, N = 3SE +/- 0.224, N = 146.2466.4997.5458.2828.76210.47712.51413.24213.39214.12715.22215.45018.54020.23421.30522.5251. (CXX) g++ options: -O3 -lm

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: SlowXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 8534PNEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9754EPYC 9654EPYC 9684XXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P1428425670SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.04, N = 3SE +/- 0.07, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.08, N = 3SE +/- 0.05, N = 4SE +/- 0.01, N = 4SE +/- 0.07, N = 4SE +/- 0.04, N = 4SE +/- 0.47, N = 8SE +/- 0.50, N = 5SE +/- 0.26, N = 517.5218.2420.2322.4623.7124.2326.0028.9631.2438.4738.9040.0040.9254.8861.0762.821. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: yolov4 - Device: CPU - Executor: StandardEPYC 9654 2PEPYC 9754 2PEPYC 9684X 2PEPYC 9754EPYC 8534PNEPYC 8534PEPYC 9684XEPYC 9654Xeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9480Xeon Platinum 8592+ 2PXeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+612182430SE +/- 0.04207, N = 3SE +/- 0.09575, N = 3SE +/- 0.08454, N = 3SE +/- 0.05314, N = 3SE +/- 0.09781, N = 5SE +/- 0.10588, N = 4SE +/- 0.12519, N = 3SE +/- 0.08554, N = 8SE +/- 0.09595, N = 14SE +/- 0.29451, N = 15SE +/- 0.06825, N = 3SE +/- 0.15263, N = 15SE +/- 0.09991, N = 3SE +/- 0.16821, N = 15SE +/- 0.04802, N = 3SE +/- 0.05084, N = 36.412296.749777.248698.295338.752338.874469.736859.8601814.4583014.4831015.5555016.9166016.9672018.2179019.8593022.968901. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Llamafile

Mozilla's Llamafile allows distributing and running large language models (LLMs) as a single file. Llamafile aims to make open-source LLMs more accessible to developers and users. Llamafile supports a variety of models, CPUs and GPUs, and other options. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.7Test: llava-v1.5-7b-q4 - Acceleration: CPUXeon Max 9480 2PXeon Platinum 8592+ 2PXeon Max 9468Xeon Max 9468 2PXeon Max 9480Xeon Platinum 8490H 2PEPYC 9754 2PEPYC 9684X 2PEPYC 9654 2PXeon Platinum 8592+EPYC 9754EPYC 8534PNXeon Platinum 8490HEPYC 8534PEPYC 9684XEPYC 9654612182430SE +/- 0.04, N = 3SE +/- 1.08, N = 12SE +/- 0.10, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.09, N = 15SE +/- 0.08, N = 4SE +/- 0.02, N = 4SE +/- 0.07, N = 4SE +/- 0.17, N = 4SE +/- 0.05, N = 4SE +/- 0.07, N = 4SE +/- 0.15, N = 4SE +/- 0.11, N = 4SE +/- 0.10, N = 5SE +/- 0.02, N = 57.267.497.677.778.1310.4410.9414.2914.4716.6120.6420.6520.9521.0025.3825.75

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.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: particle_volume/ao/real_timeXeon Max 9468EPYC 8534PNEPYC 8534PXeon Max 9480Xeon Platinum 8490HEPYC 9654Xeon Max 9468 2PXeon Platinum 8592+EPYC 9684XXeon Max 9480 2PEPYC 9754Xeon Platinum 8490H 2PEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9684X 2PEPYC 9754 2P1326395265SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.07, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 316.0516.4517.9718.5521.9328.1029.5030.8432.5233.1135.2538.2247.6154.1255.3456.69

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: particle_volume/scivis/real_timeXeon Max 9468EPYC 8534PNEPYC 8534PXeon Max 9480Xeon Platinum 8490HEPYC 9654Xeon Max 9468 2PXeon Platinum 8592+EPYC 9684XXeon Max 9480 2PEPYC 9754Xeon Platinum 8490H 2PEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9684X 2PEPYC 9754 2P1326395265SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.06, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 316.0916.4517.9418.4921.8928.0729.3530.6832.5233.1335.1538.2347.5353.9355.2656.45

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: MediumXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 8534PNEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9754EPYC 9654EPYC 9684XXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P1428425670SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.08, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 3SE +/- 0.04, N = 4SE +/- 0.07, N = 4SE +/- 0.01, N = 4SE +/- 0.27, N = 4SE +/- 0.52, N = 15SE +/- 0.56, N = 5SE +/- 0.40, N = 518.1318.8420.9723.0224.3124.8026.6929.6232.0439.0839.5840.5041.5854.5761.1163.481. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

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.

OpenBenchmarking.orgItems / Sec, More Is BetterOpenVKL 2.0.0Benchmark: vklBenchmarkCPU ISPCEPYC 8534PNXeon Max 9468EPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8592+EPYC 9654EPYC 9684XEPYC 9754Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P9001800270036004500SE +/- 8.17, N = 3SE +/- 6.66, N = 3SE +/- 1.20, N = 3SE +/- 13.96, N = 3SE +/- 1.53, N = 3SE +/- 8.62, N = 3SE +/- 4.36, N = 3SE +/- 2.08, N = 3SE +/- 1.20, N = 3SE +/- 1.00, N = 3SE +/- 0.88, N = 3SE +/- 6.74, N = 3SE +/- 13.62, N = 3SE +/- 8.67, N = 3SE +/- 5.36, N = 3SE +/- 26.33, N = 31177124312661356173417761852200420772396250827163222333539064012MIN: 91 / MAX: 14403MIN: 95 / MAX: 19286MIN: 98 / MAX: 15312MIN: 107 / MAX: 20909MIN: 104 / MAX: 26188MIN: 177 / MAX: 27371MIN: 191 / MAX: 28947MIN: 134 / MAX: 29750MIN: 160 / MAX: 26161MIN: 187 / MAX: 29055MIN: 207 / MAX: 30640MIN: 188 / MAX: 30888MIN: 248 / MAX: 36519MIN: 299 / MAX: 40960MIN: 350 / MAX: 44654MIN: 382 / MAX: 47557

Xcompact3d Incompact3d

Xcompact3d Incompact3d is a Fortran-MPI based, finite difference high-performance code for solving the incompressible Navier-Stokes equation and as many as you need scalar transport equations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: X3D-benchmarking input.i3dEPYC 8534PNEPYC 8534PEPYC 9754EPYC 9654Xeon Max 9468Xeon Platinum 8490HXeon Max 9480Xeon Platinum 8592+EPYC 9684XEPYC 9754 2PEPYC 9654 2PXeon Platinum 8490H 2PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8592+ 2PEPYC 9684X 2P120240360480600SE +/- 0.22, N = 3SE +/- 0.92, N = 3SE +/- 5.39, N = 3SE +/- 4.73, N = 3SE +/- 1.25, N = 3SE +/- 3.88, N = 4SE +/- 0.14, N = 3SE +/- 2.51, N = 3SE +/- 3.06, N = 9SE +/- 9.13, N = 9SE +/- 3.26, N = 9SE +/- 0.05, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.37, N = 3SE +/- 2.20, N = 12568.25567.29534.06466.55407.59357.01351.64323.25309.85253.32248.87196.81189.69187.22171.38167.131. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Vehicle Bike Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2PXeon Max 9480 2PXeon Max 9480Xeon Max 9468 2PEPYC 8534PNXeon Max 9468Xeon Platinum 8592+ 2PEPYC 9754EPYC 8534PXeon Platinum 8592+EPYC 9654EPYC 9684XEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P48121620SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.11, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 316.5015.0414.4213.5913.3412.6712.5411.6911.3511.2211.179.478.486.275.204.91MIN: 11.95 / MAX: 28.87MIN: 12.32 / MAX: 32.87MIN: 9.74 / MAX: 43.95MIN: 9.53 / MAX: 21.61MIN: 9.67 / MAX: 39.84MIN: 7.38 / MAX: 24.53MIN: 9.17 / MAX: 21.01MIN: 10.28 / MAX: 42.12MIN: 6.34 / MAX: 32.69MIN: 6.69 / MAX: 20.05MIN: 10.18 / MAX: 19.36MIN: 5.96 / MAX: 23.01MIN: 5 / MAX: 22.91MIN: 5.66 / MAX: 25.66MIN: 4.52 / MAX: 23.03MIN: 4.44 / MAX: 21.921. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 256 - Buffer Length: 256 - Filter Length: 32Xeon Max 9468Xeon Max 9480EPYC 8534PNEPYC 8534PXeon Platinum 8592+Xeon Platinum 8490HXeon Max 9468 2PEPYC 9654Xeon Max 9480 2PEPYC 9684XXeon Platinum 8592+ 2PXeon Platinum 8490H 2PEPYC 9754 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2P2000M4000M6000M8000M10000MSE +/- 1320353.49, N = 3SE +/- 1338323.99, N = 3SE +/- 1928730.15, N = 3SE +/- 1501110.70, N = 3SE +/- 2516611.48, N = 3SE +/- 4021608.30, N = 3SE +/- 2447674.63, N = 3SE +/- 1419311.26, N = 3SE +/- 895048.11, N = 3SE +/- 11426625.83, N = 3SE +/- 6711433.03, N = 3SE +/- 10496401.50, N = 3SE +/- 2936551.27, N = 3SE +/- 3768731.00, N = 3SE +/- 46730658.51, N = 3SE +/- 34610274.65, N = 323818000002769066667278960000030622000003167200000324380000046709666675271933333536703333358624666676077300000611656666764390000006588200000751333333379352666671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer ISPC - Model: Asian Dragon ObjEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PEPYC 9654Xeon Platinum 8490H 2PEPYC 9684XEPYC 9754Xeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9754 2PEPYC 9684X 2P50100150200250SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.08, N = 3SE +/- 0.09, N = 3SE +/- 0.09, N = 3SE +/- 0.35, N = 3SE +/- 0.76, N = 3SE +/- 0.89, N = 15SE +/- 0.02, N = 4SE +/- 0.22, N = 4SE +/- 0.11, N = 4SE +/- 0.07, N = 4SE +/- 0.29, N = 4SE +/- 1.58, N = 15SE +/- 3.26, N = 12SE +/- 1.72, N = 1563.3464.3364.9671.8379.3897.19104.84112.72113.28132.28133.48133.93173.22175.60200.16210.44MIN: 62.81 / MAX: 64.1MIN: 63.73 / MAX: 65.21MIN: 63.12 / MAX: 67.81MIN: 69.55 / MAX: 76.73MIN: 76.07 / MAX: 85.98MIN: 92.96 / MAX: 102.04MIN: 101.05 / MAX: 112.02MIN: 102 / MAX: 125.96MIN: 111.7 / MAX: 116.19MIN: 125 / MAX: 146.52MIN: 130.86 / MAX: 138.79MIN: 131.91 / MAX: 136.58MIN: 164.07 / MAX: 187.8MIN: 154.41 / MAX: 194.22MIN: 148.48 / MAX: 223.22MIN: 168.12 / MAX: 235.13

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: ThoroughXeon Max 9468EPYC 8534PNEPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PEPYC 9684XXeon Max 9480 2PEPYC 9754Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P306090120150SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 4SE +/- 0.03, N = 4SE +/- 0.01, N = 4SE +/- 0.05, N = 4SE +/- 0.01, N = 4SE +/- 0.05, N = 4SE +/- 0.06, N = 5SE +/- 0.03, N = 5SE +/- 0.01, N = 5SE +/- 0.19, N = 5SE +/- 0.55, N = 5SE +/- 0.73, N = 15SE +/- 1.46, N = 12SE +/- 1.20, N = 537.0137.9641.0242.9851.3054.4565.3072.3373.3883.3486.0186.9897.94105.41116.12121.301. (CXX) g++ options: -O3 -flto -pthread

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: SlowXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 8534PNXeon Platinum 8592+EPYC 8534PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9754EPYC 9684XEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P1020304050SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.23, N = 4SE +/- 0.21, N = 4SE +/- 0.12, N = 413.3613.8915.7115.7716.0617.0419.7522.1424.7127.1128.0129.0029.1837.1340.9643.39

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 256 - Buffer Length: 256 - Filter Length: 512EPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PXeon Max 9480 2PEPYC 9684XXeon Platinum 8490H 2PEPYC 9754 2PEPYC 9754Xeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2P500M1000M1500M2000M2500MSE +/- 170391.71, N = 3SE +/- 776881.23, N = 3SE +/- 569766.03, N = 3SE +/- 740202.67, N = 3SE +/- 1722723.94, N = 3SE +/- 240370.09, N = 3SE +/- 520683.31, N = 3SE +/- 4238841.56, N = 3SE +/- 6728628.72, N = 3SE +/- 2676440.42, N = 3SE +/- 5630373.98, N = 3SE +/- 1650252.51, N = 3SE +/- 1300000.00, N = 3SE +/- 14465476.14, N = 3SE +/- 11753202.87, N = 3SE +/- 14870701.99, N = 37417400008210133338385300009445500001020833333117876666713621666671444066667151893333315704000001616966667173480000017447000002027000000212953333323943333331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer ISPC - Model: Asian DragonEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PEPYC 9654Xeon Platinum 8490H 2PEPYC 9684XEPYC 9754EPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9684X 2P50100150200250SE +/- 0.05, N = 5SE +/- 0.03, N = 5SE +/- 0.04, N = 5SE +/- 0.11, N = 5SE +/- 0.05, N = 5SE +/- 0.13, N = 6SE +/- 0.63, N = 6SE +/- 1.20, N = 15SE +/- 0.03, N = 6SE +/- 0.13, N = 7SE +/- 0.07, N = 7SE +/- 0.12, N = 7SE +/- 1.54, N = 8SE +/- 0.26, N = 8SE +/- 1.88, N = 7SE +/- 2.04, N = 873.9175.2175.4783.4991.99112.66125.75129.75132.49152.75155.12156.39185.71199.61208.31231.43MIN: 73.25 / MAX: 75MIN: 74.55 / MAX: 76.96MIN: 72.93 / MAX: 78.77MIN: 80.58 / MAX: 88.97MIN: 88.57 / MAX: 98.54MIN: 107.88 / MAX: 117.19MIN: 120.06 / MAX: 135.13MIN: 110.67 / MAX: 151.91MIN: 130.77 / MAX: 135.13MIN: 143.8 / MAX: 168.98MIN: 152.39 / MAX: 161.68MIN: 153.92 / MAX: 160MIN: 154.13 / MAX: 227.12MIN: 185.98 / MAX: 216.64MIN: 162.41 / MAX: 258.13MIN: 187.58 / MAX: 274.64

High Performance Conjugate Gradient

HPCG is the High Performance Conjugate Gradient and is a new scientific benchmark from Sandia National Lans focused for super-computer testing with modern real-world workloads compared to HPCC. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHigh Performance Conjugate Gradient 3.1X Y Z: 144 144 144 - RT: 60EPYC 8534PNEPYC 8534PEPYC 9754Xeon Max 9468EPYC 9684XEPYC 9654Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9754 2PEPYC 9654 2PEPYC 9684X 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9480 2PXeon Platinum 8592+ 2P1530456075SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.08, N = 3SE +/- 0.28, N = 5SE +/- 0.05, N = 3SE +/- 0.37, N = 3SE +/- 0.02, N = 3SE +/- 0.08, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.51, N = 3SE +/- 0.29, N = 3SE +/- 0.03, N = 3SE +/- 0.07, N = 3SE +/- 0.94, N = 322.1322.2023.6325.2427.5828.3230.5931.0735.1441.8742.5244.9554.4160.8562.6469.301. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -lmpi_cxx -lmpi

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: GPT-2 - Device: CPU - Executor: StandardEPYC 9754 2PEPYC 8534PNEPYC 8534PEPYC 9754EPYC 9654 2PEPYC 9654Xeon Max 9480 2PXeon Max 9468 2PEPYC 9684X 2PEPYC 9684XXeon Max 9480Xeon Max 9468Xeon Platinum 8490H 2PXeon Platinum 8490HXeon Platinum 8592+ 2PXeon Platinum 8592+60120180240300SE +/- 1.43, N = 15SE +/- 0.63, N = 3SE +/- 0.28, N = 3SE +/- 0.15, N = 3SE +/- 0.13, N = 3SE +/- 1.37, N = 5SE +/- 1.08, N = 3SE +/- 0.58, N = 3SE +/- 0.46, N = 3SE +/- 0.70, N = 3SE +/- 1.88, N = 3SE +/- 2.01, N = 15SE +/- 1.05, N = 3SE +/- 1.50, N = 3SE +/- 8.14, N = 15SE +/- 1.19, N = 386.6988.5789.2189.74102.57126.82132.71141.95161.95162.09168.86176.33180.03222.03236.65270.331. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: CaffeNet 12-int8 - Device: CPU - Executor: ParallelEPYC 9754 2PEPYC 9684X 2PEPYC 9654 2PEPYC 9754Xeon Max 9480 2PXeon Max 9468 2PEPYC 8534PEPYC 8534PNXeon Platinum 8490H 2PEPYC 9654EPYC 9684XXeon Platinum 8592+ 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+2004006008001000SE +/- 2.41, N = 15SE +/- 4.10, N = 3SE +/- 2.35, N = 15SE +/- 3.02, N = 15SE +/- 1.98, N = 3SE +/- 1.77, N = 3SE +/- 0.91, N = 3SE +/- 5.34, N = 6SE +/- 4.81, N = 3SE +/- 4.99, N = 15SE +/- 5.53, N = 15SE +/- 2.57, N = 3SE +/- 4.87, N = 3SE +/- 4.16, N = 3SE +/- 5.72, N = 3SE +/- 10.06, N = 4270.63310.59315.36452.32478.47492.89547.52553.47566.38586.18590.71605.59678.48687.67784.30830.731. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Intel Open Image Denoise

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

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.2Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-OnlyEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480EPYC 9684XXeon Platinum 8490HEPYC 9654EPYC 9754Xeon Platinum 8592+EPYC 9684X 2PEPYC 9654 2PXeon Max 9468 2PEPYC 9754 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2P1.14532.29063.43594.58125.7265SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 4SE +/- 0.00, N = 4SE +/- 0.02, N = 4SE +/- 0.00, N = 4SE +/- 0.01, N = 4SE +/- 0.03, N = 6SE +/- 0.00, N = 5SE +/- 0.00, N = 5SE +/- 0.01, N = 5SE +/- 0.02, N = 5SE +/- 0.01, N = 5SE +/- 0.02, N = 5SE +/- 0.01, N = 6SE +/- 0.03, N = 61.691.802.272.392.662.712.733.163.323.473.493.493.573.664.485.09

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: super-resolution-10 - Device: CPU - Executor: ParallelEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754EPYC 8534PNEPYC 8534PEPYC 9684XEPYC 9654Xeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+ 2PXeon Platinum 8592+50100150200250SE +/- 0.83, N = 3SE +/- 0.53, N = 3SE +/- 0.41, N = 3SE +/- 0.26, N = 3SE +/- 0.72, N = 3SE +/- 0.54, N = 3SE +/- 0.99, N = 3SE +/- 0.36, N = 3SE +/- 0.62, N = 3SE +/- 1.17, N = 3SE +/- 0.74, N = 3SE +/- 0.61, N = 3SE +/- 0.98, N = 3SE +/- 2.34, N = 3SE +/- 1.38, N = 3SE +/- 1.98, N = 378.0786.9287.2694.7496.1797.05108.87110.00169.85173.91188.89190.06195.71205.53214.81234.541. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Intel Open Image Denoise

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

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.2Run: RTLightmap.hdr.4096x4096 - Device: CPU-OnlyEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480EPYC 9684XXeon Platinum 8490HEPYC 9654EPYC 9754Xeon Platinum 8592+EPYC 9684X 2PEPYC 9654 2PEPYC 9754 2PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2P0.541.081.622.162.7SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 4SE +/- 0.01, N = 40.800.861.081.141.281.291.311.501.581.641.671.711.721.742.182.40

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Noise Suppression Poconet-Like FP16 - Device: CPUEPYC 9754 2PEPYC 9654 2PEPYC 9754EPYC 9684X 2PXeon Max 9480 2PEPYC 8534PNEPYC 8534PXeon Max 9468 2PEPYC 9654Xeon Platinum 8490H 2PXeon Platinum 8490HEPYC 9684XXeon Platinum 8592+ 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8592+714212835SE +/- 0.10, N = 3SE +/- 0.08, N = 3SE +/- 0.45, N = 12SE +/- 0.04, N = 3SE +/- 0.09, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.12, N = 3SE +/- 0.16, N = 14SE +/- 0.05, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 328.1023.1222.6722.2018.7218.4917.3017.1516.3215.9613.9513.5213.1512.7211.379.40MIN: 12.27 / MAX: 100.22MIN: 10.62 / MAX: 67.83MIN: 8.31 / MAX: 66.6MIN: 9.55 / MAX: 75.9MIN: 12.15 / MAX: 64.81MIN: 11.59 / MAX: 39.13MIN: 10.72 / MAX: 37.75MIN: 10.81 / MAX: 59.34MIN: 8.11 / MAX: 42.33MIN: 10.38 / MAX: 43.64MIN: 8.67 / MAX: 65.92MIN: 6.46 / MAX: 38.94MIN: 7.28 / MAX: 47.97MIN: 9.56 / MAX: 30.66MIN: 9.17 / MAX: 32.12MIN: 7.66 / MAX: 25.071. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: MediumXeon Max 9468Xeon Max 9480Xeon Platinum 8592+EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9754EPYC 9684XEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P1020304050SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.16, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.28, N = 11SE +/- 0.23, N = 4SE +/- 0.46, N = 415.1715.6816.6417.8017.8519.2522.2524.6527.6530.3731.4432.4432.7438.1242.0345.26

Intel Open Image Denoise

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

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.2Run: RT.hdr_alb_nrm.3840x2160 - Device: CPU-OnlyEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480EPYC 9684XXeon Platinum 8490HEPYC 9654EPYC 9754Xeon Platinum 8592+EPYC 9684X 2PEPYC 9654 2PXeon Max 9468 2PEPYC 9754 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2P1.13182.26363.39544.52725.659SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 4SE +/- 0.00, N = 4SE +/- 0.00, N = 4SE +/- 0.00, N = 4SE +/- 0.01, N = 4SE +/- 0.02, N = 5SE +/- 0.00, N = 5SE +/- 0.00, N = 5SE +/- 0.01, N = 5SE +/- 0.01, N = 5SE +/- 0.01, N = 5SE +/- 0.04, N = 5SE +/- 0.03, N = 6SE +/- 0.03, N = 61.691.792.262.382.642.712.733.123.313.453.483.513.593.654.415.03

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: SharpenEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PEPYC 9684XEPYC 9754Xeon Max 9480 2PEPYC 9654 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9684X 2PEPYC 9754 2P70140210280350SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 1.20, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 1.20, N = 3SE +/- 1.67, N = 3SE +/- 1.67, N = 3SE +/- 0.88, N = 3SE +/- 3.00, N = 31051121141301571671731992012072272622652842973101. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Face Detection FP16-INT8 - Device: CPUEPYC 8534PNEPYC 8534PEPYC 9754EPYC 9654EPYC 9684XXeon Platinum 8490HXeon Platinum 8490H 2PXeon Max 9480 2PEPYC 9754 2PXeon Max 9480Xeon Max 9468 2PEPYC 9654 2PXeon Max 9468Xeon Platinum 8592+ 2PEPYC 9684X 2PXeon Platinum 8592+140280420560700SE +/- 0.12, N = 3SE +/- 0.18, N = 3SE +/- 0.18, N = 3SE +/- 0.07, N = 3SE +/- 0.43, N = 3SE +/- 0.28, N = 3SE +/- 0.00, N = 3SE +/- 0.35, N = 3SE +/- 0.03, N = 3SE +/- 0.47, N = 3SE +/- 0.40, N = 3SE +/- 0.02, N = 3SE +/- 0.15, N = 3SE +/- 0.33, N = 3SE +/- 0.43, N = 3SE +/- 0.69, N = 3654.83592.22551.07510.10471.98356.38283.11278.52275.53257.50256.63253.63236.52234.24233.63224.64MIN: 592.1 / MAX: 665.73MIN: 486.17 / MAX: 612.41MIN: 513.17 / MAX: 582.25MIN: 263.11 / MAX: 527.97MIN: 439.42 / MAX: 492.74MIN: 109.11 / MAX: 395.69MIN: 203.37 / MAX: 332.08MIN: 196.58 / MAX: 401.59MIN: 260.9 / MAX: 318.85MIN: 187.62 / MAX: 295.06MIN: 191.95 / MAX: 301.21MIN: 208.02 / MAX: 272.56MIN: 172.47 / MAX: 265.77MIN: 148.68 / MAX: 448.54MIN: 208.22 / MAX: 306.45MIN: 153.29 / MAX: 553.641. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

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.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Deconvolution Batch shapes_3d - Engine: CPUEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480EPYC 9654EPYC 9684XXeon Platinum 8490HEPYC 9654 2PEPYC 9684X 2PXeon Max 9468 2PEPYC 9754Xeon Platinum 8592+Xeon Max 9480 2PEPYC 9754 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2P0.40440.80881.21321.61762.022SE +/- 0.001685, N = 9SE +/- 0.000791, N = 9SE +/- 0.000508, N = 9SE +/- 0.000927, N = 9SE +/- 0.001484, N = 9SE +/- 0.001272, N = 9SE +/- 0.000805, N = 9SE +/- 0.001921, N = 9SE +/- 0.002070, N = 9SE +/- 0.000806, N = 9SE +/- 0.001089, N = 9SE +/- 0.000392, N = 9SE +/- 0.001240, N = 9SE +/- 0.003127, N = 9SE +/- 0.001274, N = 9SE +/- 0.000902, N = 91.7974901.4983501.2640701.1964301.1548701.1192301.0774301.0094731.0019140.9450750.9436210.9071610.8643580.8296530.7361700.624250MIN: 1.65MIN: 1.37MIN: 1.21MIN: 1.16MIN: 1.04MIN: 1.02MIN: 1.06MIN: 0.94MIN: 0.92MIN: 0.9MIN: 0.92MIN: 0.87MIN: 0.81MIN: 0.65MIN: 0.66MIN: 0.581. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: CaffeNet 12-int8 - Device: CPU - Executor: StandardEPYC 9754 2PEPYC 9754EPYC 8534PEPYC 9654 2PEPYC 8534PNEPYC 9684X 2PXeon Max 9480 2PEPYC 9654Xeon Max 9468 2PEPYC 9684XXeon Platinum 8490H 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8592+ 2PXeon Platinum 8490HXeon Platinum 8592+2004006008001000SE +/- 2.36, N = 11SE +/- 2.19, N = 3SE +/- 4.84, N = 4SE +/- 4.37, N = 15SE +/- 6.13, N = 3SE +/- 7.76, N = 15SE +/- 6.21, N = 5SE +/- 9.50, N = 15SE +/- 1.83, N = 3SE +/- 6.60, N = 3SE +/- 6.50, N = 15SE +/- 7.34, N = 3SE +/- 12.91, N = 15SE +/- 8.88, N = 3SE +/- 11.46, N = 3SE +/- 2.43, N = 3316.11382.10443.06444.85446.69446.81581.77604.05604.76672.92696.98705.61728.47754.81852.33908.191. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 256 - Buffer Length: 256 - Filter Length: 57EPYC 8534PNXeon Max 9468EPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PEPYC 9654Xeon Platinum 8490H 2PEPYC 9684XXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2P2000M4000M6000M8000M10000MSE +/- 4481443.22, N = 3SE +/- 4088330.28, N = 3SE +/- 1959875.28, N = 3SE +/- 1976810.00, N = 3SE +/- 5292237.50, N = 3SE +/- 3555434.03, N = 3SE +/- 17558125.69, N = 3SE +/- 28453880.81, N = 3SE +/- 2618735.91, N = 3SE +/- 2331189.49, N = 3SE +/- 8144527.82, N = 3SE +/- 61971606.40, N = 3SE +/- 8069352.17, N = 3SE +/- 12739745.33, N = 3SE +/- 60380193.59, N = 3SE +/- 50144890.07, N = 327937000002848933333311693333332225333333500833333380683333348251333335036000000526646666753612666675811900000628430000063162666676401366667741186666780127000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: ResNet50 v1-12-int8 - Device: CPU - Executor: StandardEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9754Xeon Max 9480 2PXeon Platinum 8592+ 2PXeon Max 9468 2PXeon Platinum 8490H 2PEPYC 8534PNEPYC 8534PEPYC 9654EPYC 9684XXeon Max 9480Xeon Platinum 8592+Xeon Platinum 8490HXeon Max 946870140210280350SE +/- 1.44, N = 15SE +/- 3.84, N = 12SE +/- 3.21, N = 15SE +/- 2.12, N = 15SE +/- 1.78, N = 3SE +/- 0.40, N = 3SE +/- 2.72, N = 15SE +/- 1.01, N = 3SE +/- 2.40, N = 15SE +/- 2.24, N = 15SE +/- 4.35, N = 15SE +/- 0.95, N = 3SE +/- 3.31, N = 5SE +/- 2.52, N = 3SE +/- 1.80, N = 3SE +/- 2.40, N = 3117.62133.49141.83160.90161.89171.23174.68177.76211.19219.06225.31243.34296.80314.06315.41330.101. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Orange Juice - Acceleration: CPUXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Platinum 8592+Xeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9754EPYC 9684XEPYC 9654 2PEPYC 9754 2PEPYC 9684X 2P816243240SE +/- 0.10, N = 3SE +/- 0.04, N = 3SE +/- 0.05, N = 3SE +/- 0.07, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.28, N = 15SE +/- 0.16, N = 15SE +/- 0.12, N = 3SE +/- 0.30, N = 15SE +/- 0.18, N = 15SE +/- 0.07, N = 3SE +/- 0.28, N = 5SE +/- 0.84, N = 15SE +/- 1.12, N = 15SE +/- 1.06, N = 1511.8612.5013.0013.4613.8114.7416.6416.9118.7019.9121.3223.7624.6929.6132.2633.08MIN: 9.93 / MAX: 13.28MIN: 10.38 / MAX: 14.93MIN: 11.12 / MAX: 14.81MIN: 11.09 / MAX: 16.56MIN: 11.85 / MAX: 16.24MIN: 12.17 / MAX: 18.21MIN: 12.39 / MAX: 24.93MIN: 12.9 / MAX: 23.07MIN: 16.22 / MAX: 25.05MIN: 15 / MAX: 30.59MIN: 18.08 / MAX: 26.9MIN: 20.86 / MAX: 33.07MIN: 20.63 / MAX: 30.25MIN: 22.93 / MAX: 46.21MIN: 23.52 / MAX: 56.17MIN: 24.95 / MAX: 50.9

GPAW

GPAW is a density-functional theory (DFT) Python code based on the projector-augmented wave (PAW) method and the atomic simulation environment (ASE). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGPAW 23.6Input: Carbon NanotubeEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Max 9480 2PXeon Platinum 8592+Xeon Max 9468 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9754EPYC 9684XEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P1530456075SE +/- 0.01, N = 3SE +/- 0.07, N = 3SE +/- 0.03, N = 3SE +/- 0.08, N = 3SE +/- 0.24, N = 3SE +/- 0.27, N = 3SE +/- 0.12, N = 3SE +/- 0.08, N = 3SE +/- 0.55, N = 3SE +/- 0.10, N = 3SE +/- 0.21, N = 3SE +/- 0.14, N = 3SE +/- 0.11, N = 3SE +/- 0.18, N = 3SE +/- 0.29, N = 15SE +/- 0.30, N = 1566.7163.8054.3848.4947.7843.9143.0542.8741.5638.0534.8733.1230.0027.1525.8424.131. (CC) gcc options: -shared -lxc -lblas -lmpi

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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Noise Suppression Poconet-Like FP16 - Device: CPUEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Platinum 8490HXeon Max 9480EPYC 9754Xeon Max 9468 2PEPYC 9654Xeon Max 9480 2PEPYC 9754 2PXeon Platinum 8592+EPYC 9684XXeon Platinum 8490H 2PEPYC 9654 2PEPYC 9684X 2PXeon Platinum 8592+ 2P2K4K6K8K10KSE +/- 3.67, N = 3SE +/- 0.64, N = 3SE +/- 15.13, N = 3SE +/- 1.87, N = 3SE +/- 7.60, N = 3SE +/- 139.41, N = 12SE +/- 37.79, N = 3SE +/- 67.22, N = 14SE +/- 28.98, N = 3SE +/- 27.90, N = 3SE +/- 10.14, N = 3SE +/- 1.84, N = 3SE +/- 13.31, N = 3SE +/- 77.59, N = 3SE +/- 59.51, N = 3SE +/- 24.76, N = 33405.943636.334208.494209.104386.485310.075565.235697.515933.656513.366757.266963.317356.637810.658231.729360.891. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: super-resolution-10 - Device: CPU - Executor: StandardEPYC 9654 2PEPYC 9754 2PEPYC 8534PNEPYC 9754EPYC 8534PEPYC 9684X 2PEPYC 9654EPYC 9684XXeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+ 2PXeon Platinum 8592+60120180240300SE +/- 0.61, N = 3SE +/- 2.46, N = 15SE +/- 0.30, N = 3SE +/- 0.26, N = 3SE +/- 0.30, N = 3SE +/- 2.15, N = 12SE +/- 0.67, N = 3SE +/- 0.58, N = 3SE +/- 1.02, N = 3SE +/- 2.30, N = 4SE +/- 4.15, N = 15SE +/- 0.25, N = 3SE +/- 0.09, N = 3SE +/- 0.16, N = 3SE +/- 3.39, N = 3SE +/- 0.03, N = 395.8498.64102.97106.13106.53119.56122.68122.95199.90205.89212.92213.99217.88227.42251.31259.191. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Memcached

Memcached is a high performance, distributed memory object caching system. This Memcached test profiles makes use of memtier_benchmark for excuting this CPU/memory-focused server benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:100Xeon Platinum 8490H 2PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8592+ 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+EPYC 8534PNEPYC 8534PEPYC 9754 2PEPYC 9754EPYC 9654EPYC 9654 2PEPYC 9684XEPYC 9684X 2P2M4M6M8M10MSE +/- 50660.01, N = 15SE +/- 26891.33, N = 3SE +/- 24083.78, N = 3SE +/- 35300.21, N = 3SE +/- 52575.97, N = 3SE +/- 21428.06, N = 3SE +/- 25335.31, N = 3SE +/- 35182.67, N = 3SE +/- 41123.40, N = 3SE +/- 36366.89, N = 3SE +/- 14591.70, N = 3SE +/- 38372.96, N = 3SE +/- 29027.89, N = 3SE +/- 83562.82, N = 3SE +/- 58081.25, N = 3SE +/- 27143.58, N = 33245908.403373323.923392842.383780240.564315467.174384040.524481667.954752673.795820618.186198256.727645325.707736018.748220895.458498306.788634258.398683410.791. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Small Mesh Size - Execution TimeEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Platinum 8490HXeon Max 9480EPYC 9754Xeon Platinum 8592+EPYC 9654Xeon Max 9468 2PXeon Platinum 8490H 2PEPYC 9684XXeon Max 9480 2PEPYC 9654 2PEPYC 9754 2PXeon Platinum 8592+ 2PEPYC 9684X 2P132639526559.2758.1745.9542.2241.7533.8133.0732.5928.9928.5628.3327.5825.0124.5723.5622.711. (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

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUEPYC 9754EPYC 8534PNEPYC 8534PXeon Platinum 8592+EPYC 9654Xeon Platinum 8592+ 2PEPYC 9684XXeon Platinum 8490HEPYC 9754 2PXeon Max 9480EPYC 9684X 2PEPYC 9654 2PXeon Max 9468Xeon Max 9480 2PXeon Platinum 8490H 2PXeon Max 9468 2P0.16430.32860.49290.65720.8215SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 9SE +/- 0.01, N = 3SE +/- 0.00, N = 7SE +/- 0.00, N = 15SE +/- 0.00, N = 3SE +/- 0.00, N = 15SE +/- 0.00, N = 3SE +/- 0.00, N = 30.730.600.530.490.480.460.450.430.410.370.350.350.330.300.290.28MIN: 0.26 / MAX: 23.1MIN: 0.26 / MAX: 24.7MIN: 0.26 / MAX: 22.12MIN: 0.18 / MAX: 16.07MIN: 0.22 / MAX: 18.92MIN: 0.18 / MAX: 22.29MIN: 0.22 / MAX: 19.11MIN: 0.23 / MAX: 17.03MIN: 0.39 / MAX: 91.48MIN: 0.28 / MAX: 40.85MIN: 0.32 / MAX: 82.66MIN: 0.32 / MAX: 81.66MIN: 0.26 / MAX: 19.2MIN: 0.26 / MAX: 42.1MIN: 0.24 / MAX: 39.69MIN: 0.25 / MAX: 41.41. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Person Re-Identification Retail FP16 - Device: CPUXeon Platinum 8490HEPYC 8534PNEPYC 8534PEPYC 9754Xeon Max 9480 2PXeon Max 9480Xeon Platinum 8490H 2PXeon Max 9468 2PEPYC 9654Xeon Max 9468Xeon Platinum 8592+ 2PEPYC 9684XXeon Platinum 8592+EPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P3691215SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 39.799.468.598.037.977.767.727.327.237.136.956.936.724.373.983.78MIN: 4.37 / MAX: 18.18MIN: 5.55 / MAX: 22.41MIN: 5.08 / MAX: 17.25MIN: 4.21 / MAX: 29.66MIN: 7.04 / MAX: 21.51MIN: 6.86 / MAX: 12.92MIN: 7.16 / MAX: 18.31MIN: 6.78 / MAX: 18.72MIN: 3.57 / MAX: 20.82MIN: 6.64 / MAX: 12.22MIN: 6.28 / MAX: 20.42MIN: 3.55 / MAX: 50.69MIN: 6.22 / MAX: 11.15MIN: 4.13 / MAX: 27.79MIN: 3.41 / MAX: 18.88MIN: 3.43 / MAX: 19.661. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: AlexNetEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 9654EPYC 9684XXeon Platinum 8592+Xeon Max 9468 2PEPYC 9754Xeon Max 9480 2PEPYC 9654 2PEPYC 9754 2PEPYC 9684X 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2P400800120016002000SE +/- 0.57, N = 3SE +/- 0.35, N = 3SE +/- 3.89, N = 3SE +/- 5.13, N = 3SE +/- 0.51, N = 3SE +/- 0.67, N = 3SE +/- 0.49, N = 3SE +/- 1.84, N = 3SE +/- 17.88, N = 4SE +/- 4.07, N = 3SE +/- 8.55, N = 3SE +/- 12.71, N = 3SE +/- 7.97, N = 3SE +/- 23.15, N = 4SE +/- 18.52, N = 3SE +/- 19.23, N = 3796.82875.231047.961108.951278.451363.331392.911486.121604.111620.531677.621862.791876.661897.641899.642039.63

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: EnhancedXeon Max 9468EPYC 8534PNEPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PEPYC 9684XXeon Max 9480 2PEPYC 9754Xeon Platinum 8490H 2PEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9684X 2P90180270360450SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 2.60, N = 3SE +/- 0.67, N = 3SE +/- 1.00, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 1.20, N = 3SE +/- 1.15, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 31601631721812162282642693003013043423683744034081. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

LAMMPS Molecular Dynamics Simulator

LAMMPS is a classical molecular dynamics code, and an acronym for Large-scale Atomic/Molecular Massively Parallel Simulator. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 23Jun2022Model: Rhodopsin ProteinEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Platinum 8490HXeon Max 9480Xeon Platinum 8592+EPYC 9654EPYC 9684XEPYC 9754Xeon Platinum 8490H 2PXeon Max 9468 2PEPYC 9654 2PEPYC 9754 2PXeon Max 9480 2PEPYC 9684X 2PXeon Platinum 8592+ 2P1224364860SE +/- 0.30, N = 15SE +/- 0.29, N = 15SE +/- 0.05, N = 11SE +/- 0.36, N = 15SE +/- 0.04, N = 11SE +/- 0.29, N = 15SE +/- 0.14, N = 11SE +/- 0.35, N = 15SE +/- 0.29, N = 10SE +/- 0.33, N = 8SE +/- 0.11, N = 9SE +/- 0.39, N = 7SE +/- 0.24, N = 5SE +/- 0.09, N = 8SE +/- 0.43, N = 7SE +/- 0.97, N = 1521.9424.2527.6328.7230.8434.3036.1540.0040.3942.2847.5150.1251.6352.2453.7755.501. (CXX) g++ options: -O3 -lm -ldl

CloverLeaf

CloverLeaf is a Lagrangian-Eulerian hydrodynamics benchmark. This test profile currently makes use of CloverLeaf's OpenMP version. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCloverLeaf 1.3Input: clover_bm64_shortEPYC 8534PNEPYC 8534PEPYC 9754EPYC 9654Xeon Max 9480EPYC 9684XXeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+Xeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9754 2PXeon Platinum 8490H 2PEPYC 9684X 2P1326395265SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 2.07, N = 12SE +/- 2.15, N = 15SE +/- 0.17, N = 3SE +/- 1.81, N = 15SE +/- 0.15, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.40, N = 4SE +/- 0.31, N = 15SE +/- 0.45, N = 3SE +/- 1.11, N = 12SE +/- 0.30, N = 3SE +/- 0.24, N = 3SE +/- 0.23, N = 357.5457.3848.3646.5042.2242.1741.4339.4839.3832.4432.2531.9030.0527.7426.5822.831. (F9X) gfortran options: -O3 -march=native -funroll-loops -fopenmp

Timed Node.js Compilation

This test profile times how long it takes to build/compile Node.js itself from source. Node.js is a JavaScript run-time built from the Chrome V8 JavaScript engine while itself is written in C/C++. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Node.js Compilation 21.7.2Time To CompileEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Platinum 8490HXeon Max 9480Xeon Platinum 8592+EPYC 9754Xeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9480 2PEPYC 9654EPYC 9684XEPYC 9754 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9684X 2P60120180240300SE +/- 0.19, N = 3SE +/- 0.17, N = 3SE +/- 0.53, N = 3SE +/- 0.71, N = 3SE +/- 0.74, N = 3SE +/- 0.52, N = 3SE +/- 0.72, N = 3SE +/- 0.84, N = 3SE +/- 0.48, N = 3SE +/- 0.52, N = 3SE +/- 0.59, N = 3SE +/- 0.50, N = 3SE +/- 1.51, N = 3SE +/- 0.12, N = 3SE +/- 0.18, N = 3SE +/- 1.19, N = 3288.12277.53270.38254.00252.65214.43186.10175.63172.99171.17164.02147.84147.74145.41118.90114.91

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: GPT-2 - Device: CPU - Executor: ParallelEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2PEPYC 8534PEPYC 8534PNXeon Max 9468 2PXeon Max 9480 2PEPYC 9754Xeon Platinum 8490H 2PEPYC 9654EPYC 9684XXeon Max 9480Xeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+ 2PXeon Platinum 8592+50100150200250SE +/- 0.19, N = 3SE +/- 0.35, N = 3SE +/- 0.48, N = 3SE +/- 0.64, N = 3SE +/- 0.43, N = 3SE +/- 1.53, N = 3SE +/- 0.26, N = 3SE +/- 0.17, N = 3SE +/- 0.78, N = 3SE +/- 0.08, N = 3SE +/- 0.64, N = 3SE +/- 2.63, N = 3SE +/- 1.04, N = 3SE +/- 0.53, N = 3SE +/- 0.81, N = 3SE +/- 0.27, N = 399.22109.46110.17128.98129.65135.98136.69137.19144.64156.61161.50188.93191.11191.56195.45248.681. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

7-Zip Compression

This is a test of 7-Zip compression/decompression with its integrated benchmark feature. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Compression RatingXeon Max 9468EPYC 8534PNEPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PEPYC 9654EPYC 9754EPYC 9684XXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9684X 2PEPYC 9654 2PEPYC 9754 2P170K340K510K680K850KSE +/- 1836.55, N = 3SE +/- 206.69, N = 3SE +/- 268.94, N = 3SE +/- 2258.81, N = 3SE +/- 1535.53, N = 3SE +/- 1982.97, N = 3SE +/- 5127.15, N = 3SE +/- 3924.87, N = 3SE +/- 719.89, N = 3SE +/- 1188.96, N = 3SE +/- 1520.59, N = 3SE +/- 4200.62, N = 3SE +/- 709.87, N = 3SE +/- 3075.95, N = 3SE +/- 6869.20, N = 3SE +/- 2500.52, N = 33147163190943313573423633881904416364917295255145495005754405791125927296742397413667617347769681. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: bertsquad-12 - Device: CPU - Executor: ParallelEPYC 9754 2PEPYC 9684X 2PEPYC 9654 2PEPYC 8534PNEPYC 9754EPYC 8534PEPYC 9654EPYC 9684XXeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PXeon Max 9480Xeon Platinum 8490HXeon Max 9468Xeon Platinum 8592+510152025SE +/- 0.01907, N = 3SE +/- 0.04604, N = 3SE +/- 0.05052, N = 3SE +/- 0.10146, N = 4SE +/- 0.05587, N = 3SE +/- 0.10485, N = 3SE +/- 0.09004, N = 3SE +/- 0.08493, N = 3SE +/- 0.04124, N = 3SE +/- 0.09371, N = 3SE +/- 0.02421, N = 3SE +/- 0.02832, N = 3SE +/- 0.07567, N = 3SE +/- 0.04566, N = 3SE +/- 0.06023, N = 3SE +/- 0.21736, N = 37.984028.759718.809308.871449.652969.8009012.2549012.2807013.5190013.7366014.5280015.4341015.7394016.0956016.3251019.677301. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Parallel BZIP2 Compression

This test measures the time needed to compress a file (FreeBSD-13.0-RELEASE-amd64-memstick.img) using Parallel BZIP2 compression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterParallel BZIP2 Compression 1.1.13FreeBSD-13.0-RELEASE-amd64-memstick.img CompressionXeon Max 9468EPYC 8534PNXeon Max 9480EPYC 8534PXeon Platinum 8490HXeon Max 9468 2PXeon Platinum 8592+Xeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9684XEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P0.59221.18441.77662.36882.961SE +/- 0.023295, N = 15SE +/- 0.016947, N = 15SE +/- 0.020941, N = 15SE +/- 0.023387, N = 15SE +/- 0.007749, N = 10SE +/- 0.030907, N = 15SE +/- 0.023842, N = 15SE +/- 0.028863, N = 15SE +/- 0.017240, N = 15SE +/- 0.024402, N = 15SE +/- 0.014512, N = 15SE +/- 0.015384, N = 15SE +/- 0.010826, N = 15SE +/- 0.019468, N = 15SE +/- 0.015217, N = 15SE +/- 0.012053, N = 152.6320772.3211952.2877662.2004972.1448571.9829321.8475371.7112961.4692101.4323621.4229541.3690631.3557371.1263291.1094461.0683981. (CXX) g++ options: -O2 -pthread -lbz2 -lpthread

LAMMPS Molecular Dynamics Simulator

LAMMPS is a classical molecular dynamics code, and an acronym for Large-scale Atomic/Molecular Massively Parallel Simulator. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 23Jun2022Model: 20k AtomsEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PEPYC 9684XEPYC 9754Xeon Platinum 8490H 2PXeon Max 9480 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9754 2PEPYC 9684X 2P1428425670SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.12, N = 3SE +/- 0.07, N = 3SE +/- 0.30, N = 3SE +/- 0.19, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.04, N = 3SE +/- 0.13, N = 3SE +/- 0.13, N = 3SE +/- 0.12, N = 325.3327.5328.2832.1634.8937.4642.5146.7146.8947.9950.5251.8256.4057.1058.8462.001. (CXX) g++ options: -O3 -lm -ldl

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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUEPYC 8534PNXeon Platinum 8592+EPYC 8534PXeon Platinum 8490HXeon Platinum 8592+ 2PEPYC 9754EPYC 9654Xeon Max 9468EPYC 9684XXeon Max 9480EPYC 9754 2PEPYC 9654 2PXeon Max 9468 2PXeon Max 9480 2PEPYC 9684X 2PXeon Platinum 8490H 2P40K80K120K160K200KSE +/- 55.52, N = 3SE +/- 988.12, N = 3SE +/- 86.99, N = 3SE +/- 75.30, N = 3SE +/- 1158.60, N = 3SE +/- 89.35, N = 3SE +/- 205.87, N = 3SE +/- 383.22, N = 3SE +/- 162.47, N = 3SE +/- 1761.33, N = 3SE +/- 1374.23, N = 9SE +/- 1922.38, N = 15SE +/- 18.38, N = 3SE +/- 1422.88, N = 15SE +/- 1640.61, N = 7SE +/- 2494.14, N = 378334.9281279.7987511.11115106.30128566.81132973.43133779.62135075.07138605.52140144.01166151.42172688.50174862.56176900.32178488.03189893.311. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Timed LLVM Compilation

This test times how long it takes to compile/build the LLVM compiler stack. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: NinjaEPYC 8534PNXeon Max 9468EPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9754EPYC 9654Xeon Max 9468 2PXeon Max 9480 2PEPYC 9684XXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P50100150200250SE +/- 0.10, N = 3SE +/- 0.19, N = 3SE +/- 0.20, N = 3SE +/- 0.36, N = 3SE +/- 0.07, N = 3SE +/- 0.12, N = 3SE +/- 0.74, N = 3SE +/- 0.31, N = 3SE +/- 0.17, N = 3SE +/- 0.79, N = 3SE +/- 0.27, N = 3SE +/- 0.15, N = 3SE +/- 0.27, N = 3SE +/- 0.45, N = 3SE +/- 0.08, N = 3SE +/- 0.09, N = 3215.93212.48205.01197.84187.02165.05146.76135.67134.42128.07124.46122.31106.82101.5694.8189.28

PostgreSQL

This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 1000 - Mode: Read Write - Average LatencyXeon Platinum 8592+ 2PEPYC 9684X 2PEPYC 9654 2PEPYC 8534PNXeon Max 9468 2PEPYC 9754 2PXeon Max 9480 2PEPYC 8534PXeon Platinum 8490H 2PXeon Max 9468Xeon Platinum 8490HXeon Max 9480EPYC 9754Xeon Platinum 8592+EPYC 9684XEPYC 9654714212835SE +/- 0.10, N = 3SE +/- 0.32, N = 12SE +/- 0.25, N = 12SE +/- 0.10, N = 3SE +/- 0.06, N = 3SE +/- 0.02, N = 3SE +/- 0.12, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 330.2018.1618.0618.0516.9616.8816.6616.3214.8614.2914.0413.8513.8413.2612.5812.501. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 1000 - Mode: Read WriteXeon Platinum 8592+ 2PEPYC 9684X 2PEPYC 8534PNEPYC 9654 2PXeon Max 9468 2PEPYC 9754 2PXeon Max 9480 2PEPYC 8534PXeon Platinum 8490H 2PXeon Max 9468Xeon Platinum 8490HXeon Max 9480EPYC 9754Xeon Platinum 8592+EPYC 9684XEPYC 965420K40K60K80K100KSE +/- 106.16, N = 3SE +/- 1059.49, N = 12SE +/- 299.71, N = 3SE +/- 865.71, N = 12SE +/- 195.77, N = 3SE +/- 58.85, N = 3SE +/- 440.63, N = 3SE +/- 175.33, N = 3SE +/- 137.82, N = 3SE +/- 95.84, N = 3SE +/- 78.02, N = 3SE +/- 143.25, N = 3SE +/- 242.45, N = 3SE +/- 198.41, N = 3SE +/- 87.23, N = 3SE +/- 232.97, N = 3331105527255395555055895259228600466127267315699697122772225722447539679492799871. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: allmodconfigEPYC 8534PNXeon Max 9468EPYC 8534PXeon Max 9480Xeon Platinum 8490HEPYC 9754Xeon Platinum 8592+EPYC 9654EPYC 9684XXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9754 2PEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9684X 2P90180270360450SE +/- 0.78, N = 3SE +/- 0.61, N = 3SE +/- 0.91, N = 3SE +/- 0.67, N = 3SE +/- 0.80, N = 3SE +/- 1.05, N = 3SE +/- 0.72, N = 3SE +/- 0.49, N = 3SE +/- 0.37, N = 3SE +/- 0.56, N = 3SE +/- 0.73, N = 3SE +/- 0.86, N = 3SE +/- 1.31, N = 3SE +/- 0.49, N = 3SE +/- 0.56, N = 3SE +/- 0.98, N = 3438.17428.39406.41384.06362.45337.69316.90311.83290.34246.94231.03218.57210.20197.63189.74182.00

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: bertsquad-12 - Device: CPU - Executor: StandardEPYC 9654 2PEPYC 9754 2PEPYC 9684X 2PEPYC 9754EPYC 8534PNEPYC 8534PEPYC 9654EPYC 9684XXeon Platinum 8490H 2PXeon Max 9468 2PXeon Max 9480 2PXeon Max 9480Xeon Platinum 8490HXeon Max 9468Xeon Platinum 8592+ 2PXeon Platinum 8592+612182430SE +/- 0.10, N = 6SE +/- 0.27, N = 12SE +/- 0.39, N = 15SE +/- 0.09, N = 15SE +/- 0.17, N = 3SE +/- 0.08, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.17, N = 3SE +/- 0.19, N = 3SE +/- 0.35, N = 15SE +/- 0.09, N = 3SE +/- 0.07, N = 3SE +/- 0.20, N = 3SE +/- 0.22, N = 3SE +/- 0.02, N = 310.6610.7511.2711.4211.5412.3913.8613.9716.5716.6417.9920.0120.1521.1722.4325.161. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: INIVOL and Fluid Structure Interaction Drop ContainerEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9754EPYC 9654Xeon Platinum 8592+ 2PEPYC 9684XEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P4080120160200SE +/- 0.63, N = 3SE +/- 0.37, N = 3SE +/- 0.08, N = 3SE +/- 0.06, N = 3SE +/- 0.19, N = 3SE +/- 0.12, N = 3SE +/- 0.07, N = 3SE +/- 0.09, N = 3SE +/- 0.17, N = 3SE +/- 0.46, N = 3SE +/- 0.57, N = 3SE +/- 0.30, N = 3SE +/- 0.93, N = 3SE +/- 0.12, N = 3SE +/- 0.14, N = 3SE +/- 0.43, N = 3187.03176.87160.56146.03140.43128.74114.29110.56107.36106.54100.24100.0094.7690.8585.8979.42

CloverLeaf

CloverLeaf is a Lagrangian-Eulerian hydrodynamics benchmark. This test profile currently makes use of CloverLeaf's OpenMP version. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCloverLeaf 1.3Input: clover_bm16EPYC 8534PNEPYC 8534PXeon Platinum 8592+EPYC 9754Xeon Platinum 8592+ 2PXeon Max 9480Xeon Max 9468EPYC 9754 2PEPYC 9654Xeon Platinum 8490HXeon Max 9480 2PEPYC 9654 2PXeon Max 9468 2PEPYC 9684XEPYC 9684X 2PXeon Platinum 8490H 2P110220330440550SE +/- 1.08, N = 3SE +/- 0.79, N = 3SE +/- 0.54, N = 3SE +/- 4.99, N = 9SE +/- 2.02, N = 3SE +/- 2.03, N = 3SE +/- 1.43, N = 3SE +/- 1.02, N = 3SE +/- 3.79, N = 9SE +/- 0.13, N = 3SE +/- 2.11, N = 9SE +/- 0.12, N = 3SE +/- 1.56, N = 3SE +/- 2.24, N = 3SE +/- 0.59, N = 3SE +/- 0.74, N = 3498.22495.25410.59400.52378.05357.57351.29349.67334.30328.47269.20266.10263.33245.82223.54214.801. (F9X) gfortran options: -O3 -march=native -funroll-loops -fopenmp

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: AlexNetEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654EPYC 9684XEPYC 9754 2PEPYC 9754Xeon Max 9468 2PEPYC 9654 2PXeon Max 9480 2PEPYC 9684X 2PXeon Platinum 8592+ 2PXeon Platinum 8490H 2P400800120016002000SE +/- 0.41, N = 3SE +/- 0.99, N = 3SE +/- 1.21, N = 3SE +/- 6.82, N = 3SE +/- 0.54, N = 3SE +/- 4.27, N = 3SE +/- 2.58, N = 3SE +/- 1.54, N = 3SE +/- 10.23, N = 15SE +/- 4.01, N = 3SE +/- 1.15, N = 3SE +/- 6.01, N = 3SE +/- 5.81, N = 3SE +/- 15.82, N = 3SE +/- 1.86, N = 3SE +/- 5.06, N = 3771.47843.63986.041058.991205.301277.461278.581301.261403.461454.711479.531524.741528.411560.561681.441730.25

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: MediumXeon Max 9468EPYC 8534PNXeon Max 9480EPYC 8534PXeon Platinum 8592+Xeon Platinum 8490HXeon Max 9468 2PEPYC 9654EPYC 9754 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654 2PEPYC 9684XEPYC 9684X 2PEPYC 9754Xeon Platinum 8592+ 2P120240360480600SE +/- 0.03, N = 6SE +/- 0.07, N = 6SE +/- 0.08, N = 6SE +/- 0.07, N = 6SE +/- 1.17, N = 6SE +/- 0.16, N = 6SE +/- 0.52, N = 6SE +/- 0.20, N = 7SE +/- 0.52, N = 6SE +/- 0.77, N = 6SE +/- 2.79, N = 6SE +/- 3.54, N = 6SE +/- 0.63, N = 7SE +/- 3.11, N = 7SE +/- 0.39, N = 7SE +/- 1.56, N = 7247.30273.67284.33294.62327.78333.99441.65449.11462.24483.10484.55491.53500.77508.09527.80550.021. (CXX) g++ options: -O3 -flto -pthread

Y-Cruncher

Y-Cruncher is a multi-threaded Pi benchmark capable of computing Pi to trillions of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.3Pi Digits To Calculate: 500MEPYC 8534PNEPYC 8534PEPYC 9684X 2PEPYC 9654 2PEPYC 9684XEPYC 9754 2PEPYC 9754EPYC 9654Xeon Max 9468Xeon Max 9480Xeon Platinum 8592+Xeon Platinum 8490HXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8592+ 2PXeon Platinum 8490H 2P1.16212.32423.48634.64845.8105SE +/- 0.008, N = 6SE +/- 0.006, N = 6SE +/- 0.007, N = 5SE +/- 0.003, N = 5SE +/- 0.003, N = 6SE +/- 0.015, N = 5SE +/- 0.005, N = 6SE +/- 0.003, N = 6SE +/- 0.010, N = 7SE +/- 0.023, N = 15SE +/- 0.008, N = 7SE +/- 0.003, N = 7SE +/- 0.064, N = 15SE +/- 0.074, N = 15SE +/- 0.004, N = 7SE +/- 0.002, N = 85.1655.0774.5104.3294.3124.1003.9683.8503.7813.7043.3913.2673.1102.9612.6592.364

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: GoogLeNetEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2PEPYC 8534PNEPYC 8534PXeon Platinum 8592+ 2PXeon Max 9480 2PEPYC 9754Xeon Platinum 8592+EPYC 9684XXeon Platinum 8490H 2PXeon Max 9468Xeon Max 9480Xeon Max 9468 2PEPYC 9654Xeon Platinum 8490H80160240320400SE +/- 1.18, N = 3SE +/- 2.41, N = 3SE +/- 2.47, N = 3SE +/- 0.44, N = 3SE +/- 0.30, N = 3SE +/- 0.21, N = 3SE +/- 2.38, N = 3SE +/- 3.84, N = 4SE +/- 1.12, N = 3SE +/- 1.66, N = 3SE +/- 2.37, N = 3SE +/- 1.94, N = 3SE +/- 0.83, N = 3SE +/- 3.52, N = 5SE +/- 4.22, N = 3SE +/- 1.62, N = 3176.00227.22248.03256.79276.93291.37308.95314.67319.76330.26331.76333.73335.58341.51354.59380.14

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.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Handwritten English Recognition FP16-INT8 - Device: CPUEPYC 8534PNEPYC 8534PEPYC 9754 2PEPYC 9754EPYC 9654 2PXeon Platinum 8490HEPYC 9654EPYC 9684X 2PEPYC 9684XXeon Max 9480 2PXeon Platinum 8490H 2PXeon Max 9480Xeon Max 9468 2PXeon Max 9468Xeon Platinum 8592+ 2PXeon Platinum 8592+1020304050SE +/- 0.02, N = 3SE +/- 0.09, N = 3SE +/- 0.08, N = 3SE +/- 0.09, N = 3SE +/- 0.07, N = 3SE +/- 0.04, N = 3SE +/- 0.06, N = 3SE +/- 0.07, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 345.7141.3839.8638.8536.9036.2034.9233.1832.0025.7225.1224.9424.0023.1222.9621.87MIN: 31.06 / MAX: 64.63MIN: 25.29 / MAX: 59.37MIN: 35.87 / MAX: 58.79MIN: 23.02 / MAX: 69.97MIN: 29.34 / MAX: 49.06MIN: 20.73 / MAX: 45.28MIN: 22.04 / MAX: 54.74MIN: 29.34 / MAX: 51.4MIN: 21.72 / MAX: 52.65MIN: 20.7 / MAX: 44.98MIN: 19.86 / MAX: 59.96MIN: 19.75 / MAX: 33.18MIN: 19.19 / MAX: 45.88MIN: 18.74 / MAX: 31.08MIN: 20.8 / MAX: 34.86MIN: 20.16 / MAX: 32.361. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Y-Cruncher

Y-Cruncher is a multi-threaded Pi benchmark capable of computing Pi to trillions of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.3Pi Digits To Calculate: 1BEPYC 8534PNEPYC 8534PEPYC 9684X 2PXeon Max 9468EPYC 9654 2PXeon Max 9480EPYC 9754 2PEPYC 9684XXeon Platinum 8490HEPYC 9754EPYC 9654Xeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8592+ 2PXeon Platinum 8490H 2P3691215SE +/- 0.001, N = 4SE +/- 0.005, N = 4SE +/- 0.012, N = 4SE +/- 0.045, N = 5SE +/- 0.012, N = 4SE +/- 0.026, N = 5SE +/- 0.049, N = 4SE +/- 0.008, N = 4SE +/- 0.005, N = 5SE +/- 0.007, N = 4SE +/- 0.008, N = 5SE +/- 0.030, N = 5SE +/- 0.122, N = 15SE +/- 0.142, N = 12SE +/- 0.010, N = 6SE +/- 0.007, N = 610.42710.2058.9928.3008.1967.9597.9157.4717.0476.9286.7436.7016.3346.2665.1105.008

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Very FastXeon Platinum 8592+Xeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 8534PNXeon Max 9468 2PXeon Max 9480 2PEPYC 8534PEPYC 9754 2PXeon Platinum 8592+ 2PXeon Platinum 8490H 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9654EPYC 9684X20406080100SE +/- 0.04, N = 4SE +/- 0.06, N = 4SE +/- 0.08, N = 4SE +/- 0.08, N = 4SE +/- 0.07, N = 4SE +/- 0.33, N = 4SE +/- 0.29, N = 4SE +/- 0.07, N = 5SE +/- 0.63, N = 15SE +/- 0.36, N = 5SE +/- 0.45, N = 5SE +/- 0.21, N = 5SE +/- 0.71, N = 6SE +/- 0.28, N = 6SE +/- 0.62, N = 6SE +/- 0.32, N = 638.0240.2341.3450.0351.5151.9152.2354.6059.0165.3867.6468.5672.6874.0376.8378.151. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 128 - Buffer Length: 256 - Filter Length: 512EPYC 8534PNXeon Max 9468EPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Max 9468 2PXeon Platinum 8490H 2PXeon Platinum 8592+EPYC 9654EPYC 9654 2PXeon Max 9480 2PEPYC 9754 2PEPYC 9684XEPYC 9754EPYC 9684X 2PXeon Platinum 8592+ 2P300M600M900M1200M1500MSE +/- 431058.13, N = 3SE +/- 681281.47, N = 3SE +/- 390953.25, N = 3SE +/- 1345523.44, N = 3SE +/- 1740596.96, N = 3SE +/- 11560043.25, N = 5SE +/- 6842108.19, N = 3SE +/- 2313246.88, N = 3SE +/- 1047748.91, N = 3SE +/- 2338327.99, N = 3SE +/- 2123152.79, N = 3SE +/- 503322.30, N = 3SE +/- 2412697.89, N = 3SE +/- 814452.78, N = 3SE +/- 3901424.24, N = 3SE +/- 4790383.89, N = 3731756667800186667809916667882200000918566667109146000011264666671152933333115816666711999666671206066667134980000013687333331369800000139473333314938333331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

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.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Face Detection FP16-INT8 - Device: CPUEPYC 8534PNEPYC 8534PEPYC 9654EPYC 9684XEPYC 9754Xeon Platinum 8490HEPYC 9654 2PXeon Max 9468EPYC 9684X 2PXeon Max 9480EPYC 9754 2PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2P120240360480600SE +/- 0.06, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.12, N = 3SE +/- 0.03, N = 3SE +/- 0.12, N = 3SE +/- 0.01, N = 3SE +/- 0.14, N = 3SE +/- 0.41, N = 3SE +/- 0.41, N = 3SE +/- 0.04, N = 3SE +/- 0.69, N = 3SE +/- 0.58, N = 3SE +/- 0.54, N = 3SE +/- 0.01, N = 3SE +/- 0.77, N = 348.7353.8993.99101.51115.78167.83188.74202.59205.03217.11231.83284.64373.50401.45423.18545.211. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: GoogLeNetEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480EPYC 9684XEPYC 9754Xeon Platinum 8490HXeon Platinum 8592+EPYC 9684X 2PXeon Max 9468 2PEPYC 9654Xeon Max 9480 2PEPYC 9654 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9754 2P130260390520650SE +/- 0.03, N = 3SE +/- 0.06, N = 3SE +/- 0.41, N = 3SE +/- 0.37, N = 3SE +/- 3.05, N = 12SE +/- 12.03, N = 12SE +/- 0.34, N = 3SE +/- 3.68, N = 3SE +/- 1.60, N = 3SE +/- 1.14, N = 3SE +/- 4.02, N = 10SE +/- 2.46, N = 3SE +/- 5.11, N = 3SE +/- 2.62, N = 3SE +/- 1.52, N = 3SE +/- 5.63, N = 6295.29313.50383.57398.40425.25450.61461.23462.78464.78467.91470.95477.33525.84576.08577.76588.43

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: LuxCore Benchmark - Acceleration: CPUXeon Max 9468Xeon Max 9468 2PXeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Max 9480 2PXeon Platinum 8592+Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9654EPYC 9754EPYC 9684X3691215SE +/- 0.03, N = 3SE +/- 0.22, N = 12SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.16, N = 15SE +/- 0.11, N = 3SE +/- 0.12, N = 3SE +/- 0.17, N = 12SE +/- 0.08, N = 15SE +/- 0.11, N = 15SE +/- 0.12, N = 15SE +/- 0.10, N = 7SE +/- 0.16, N = 3SE +/- 0.04, N = 36.306.816.826.937.107.387.418.068.828.889.6510.5610.6111.3412.4112.43MIN: 2.53 / MAX: 7.3MIN: 2.55 / MAX: 9.19MIN: 2.81 / MAX: 7.94MIN: 2.6 / MAX: 8.15MIN: 3.06 / MAX: 8.2MIN: 2.8 / MAX: 8.68MIN: 2.67 / MAX: 9.42MIN: 3.62 / MAX: 9.37MIN: 3.72 / MAX: 10.41MIN: 3.73 / MAX: 11.48MIN: 3.42 / MAX: 12.1MIN: 3.72 / MAX: 13.3MIN: 3.78 / MAX: 13.37MIN: 4.84 / MAX: 13.61MIN: 5.2 / MAX: 14.7MIN: 5.47 / MAX: 14.43

GNU Octave Benchmark

This test profile measures how long it takes to complete several reference GNU Octave files via octave-benchmark. GNU Octave is used for numerical computations and is an open-source alternative to MATLAB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGNU Octave Benchmark 8.4.0Xeon Max 9468 2PXeon Max 9480 2PEPYC 8534PNEPYC 8534PEPYC 9754 2PEPYC 9754Xeon Platinum 8490H 2PXeon Max 9480EPYC 9654 2PXeon Max 9468Xeon Platinum 8490HEPYC 9654EPYC 9684X 2PEPYC 9684XXeon Platinum 8592+ 2PXeon Platinum 8592+3691215SE +/- 0.096, N = 5SE +/- 0.082, N = 5SE +/- 0.007, N = 5SE +/- 0.011, N = 5SE +/- 0.025, N = 5SE +/- 0.014, N = 5SE +/- 0.019, N = 6SE +/- 0.024, N = 6SE +/- 0.004, N = 6SE +/- 0.015, N = 6SE +/- 0.015, N = 6SE +/- 0.009, N = 6SE +/- 0.007, N = 6SE +/- 0.018, N = 6SE +/- 0.019, N = 6SE +/- 0.020, N = 612.61712.5788.9488.9298.8488.8018.0367.7487.6457.6227.5327.5286.9556.8576.8566.458

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: SwirlEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Max 9468 2PEPYC 9654Xeon Platinum 8592+EPYC 9754EPYC 9684XEPYC 9654 2PXeon Max 9480 2PEPYC 9754 2PEPYC 9684X 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2P150300450600750SE +/- 0.88, N = 3SE +/- 0.33, N = 3SE +/- 0.58, N = 3SE +/- 0.67, N = 3SE +/- 1.15, N = 3SE +/- 1.33, N = 3SE +/- 0.33, N = 3SE +/- 0.88, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 1.76, N = 3SE +/- 0.58, N = 33633793944325025095115295355465535585675836176891. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: T5 Encoder - Device: CPU - Executor: ParallelXeon Platinum 8490H 2PXeon Max 9468 2PEPYC 9754 2PXeon Max 9480 2PEPYC 8534PNEPYC 8534PEPYC 9654 2PEPYC 9684X 2PEPYC 9754Xeon Platinum 8490HXeon Max 9468Xeon Max 9480Xeon Platinum 8592+ 2PEPYC 9654EPYC 9684XXeon Platinum 8592+100200300400500SE +/- 2.94, N = 3SE +/- 2.44, N = 3SE +/- 0.64, N = 3SE +/- 1.67, N = 15SE +/- 1.96, N = 3SE +/- 2.24, N = 3SE +/- 2.43, N = 3SE +/- 0.65, N = 3SE +/- 3.02, N = 3SE +/- 0.86, N = 3SE +/- 3.48, N = 15SE +/- 1.74, N = 3SE +/- 1.73, N = 3SE +/- 0.55, N = 3SE +/- 3.58, N = 3SE +/- 3.84, N = 3246.34246.68250.63252.67254.73255.67271.62276.21319.45326.14337.00346.21354.75374.75446.30462.861. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 128 - Buffer Length: 256 - Filter Length: 32Xeon Max 9468Xeon Max 9480EPYC 8534PNEPYC 8534PXeon Platinum 8592+Xeon Platinum 8490HXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8592+ 2PXeon Platinum 8490H 2PEPYC 9754EPYC 9754 2PEPYC 9654EPYC 9654 2PEPYC 9684XEPYC 9684X 2P900M1800M2700M3600M4500MSE +/- 400000.00, N = 3SE +/- 1342054.81, N = 3SE +/- 2003607.86, N = 3SE +/- 3138116.92, N = 3SE +/- 2236316.42, N = 3SE +/- 1858314.65, N = 3SE +/- 4598671.31, N = 3SE +/- 2403700.85, N = 3SE +/- 13813198.20, N = 3SE +/- 26633124.74, N = 3SE +/- 4056818.68, N = 3SE +/- 384418.75, N = 3SE +/- 2714365.08, N = 3SE +/- 3186603.90, N = 3SE +/- 841295.30, N = 3SE +/- 4250882.26, N = 323519000002739633333276793333330470666673148233333320060000032540333333337066667352076666735586000003791566667379413333340327333334063333333435353333344079000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: GoogLeNetEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480EPYC 9684X 2PEPYC 9754 2PEPYC 9654 2PEPYC 9684XXeon Platinum 8592+Xeon Max 9468 2PEPYC 9654Xeon Platinum 8490HXeon Max 9480 2PEPYC 9754Xeon Platinum 8592+ 2PXeon Platinum 8490H 2P110220330440550SE +/- 0.07, N = 3SE +/- 0.04, N = 3SE +/- 1.67, N = 3SE +/- 1.91, N = 3SE +/- 0.51, N = 3SE +/- 1.11, N = 3SE +/- 1.02, N = 3SE +/- 2.76, N = 15SE +/- 2.30, N = 3SE +/- 2.29, N = 3SE +/- 3.87, N = 15SE +/- 0.24, N = 3SE +/- 4.80, N = 3SE +/- 10.93, N = 12SE +/- 2.68, N = 3SE +/- 4.87, N = 3287.42308.46375.73386.00391.40397.44401.84406.58423.59447.50447.61448.46449.91462.79501.39528.93

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: ResNet-50EPYC 8534PNEPYC 8534PXeon Max 9468EPYC 9684XXeon Max 9480EPYC 9754EPYC 9684X 2PXeon Max 9480 2PEPYC 9654Xeon Platinum 8592+Xeon Max 9468 2PXeon Platinum 8490HEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9754 2PXeon Platinum 8490H 2P4080120160200SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 1.17, N = 3SE +/- 1.42, N = 4SE +/- 0.61, N = 3SE +/- 1.62, N = 4SE +/- 0.31, N = 3SE +/- 1.80, N = 3SE +/- 0.34, N = 3SE +/- 0.48, N = 3SE +/- 0.79, N = 3SE +/- 0.01, N = 3SE +/- 1.83, N = 3SE +/- 0.61, N = 3SE +/- 0.81, N = 3SE +/- 1.39, N = 395.63101.17117.53121.96125.14135.17137.50138.39140.01140.53145.29148.45161.90169.75174.93174.95

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100Xeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9754 2PEPYC 8534PEPYC 8534PNEPYC 9684X 2PEPYC 9654 2PEPYC 9754EPYC 9654EPYC 9684X30M60M90M120M150MSE +/- 872659.60, N = 4SE +/- 594234.55, N = 3SE +/- 928557.10, N = 3SE +/- 435900.66, N = 3SE +/- 782176.98, N = 3SE +/- 276414.35, N = 3SE +/- 862582.24, N = 7SE +/- 739343.51, N = 3SE +/- 1225528.40, N = 12SE +/- 576868.09, N = 3SE +/- 41021.96, N = 3SE +/- 1058559.19, N = 12SE +/- 1249999.63, N = 12SE +/- 263772.09, N = 3SE +/- 665173.47, N = 3SE +/- 538692.55, N = 38013956680770465836517858574454286354518908717709277163994616807115317834116879799117884977124947288125595607131781309142102254144760448

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: ResNet-50EPYC 9754 2PEPYC 8534PNEPYC 9654 2PEPYC 9684X 2PEPYC 8534PXeon Platinum 8592+Xeon Platinum 8592+ 2PXeon Max 9480Xeon Max 9468EPYC 9684XEPYC 9754Xeon Max 9480 2PXeon Max 9468 2PEPYC 9654Xeon Platinum 8490H 2PXeon Platinum 8490H306090120150SE +/- 0.88, N = 3SE +/- 0.11, N = 3SE +/- 0.91, N = 5SE +/- 1.04, N = 4SE +/- 0.05, N = 3SE +/- 0.50, N = 3SE +/- 0.70, N = 3SE +/- 0.93, N = 3SE +/- 0.13, N = 3SE +/- 0.08, N = 3SE +/- 0.60, N = 3SE +/- 1.28, N = 3SE +/- 0.39, N = 3SE +/- 0.64, N = 3SE +/- 1.12, N = 3SE +/- 0.04, N = 367.0880.8682.9486.0887.3191.5194.66103.56104.23104.52106.14107.30111.71114.63115.47120.93

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100Xeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9754 2PEPYC 8534PEPYC 8534PNEPYC 9684X 2PEPYC 9654 2PEPYC 9754EPYC 9654EPYC 9684X20406080100SE +/- 1.11, N = 4SE +/- 1.01, N = 3SE +/- 1.18, N = 3SE +/- 0.48, N = 3SE +/- 0.87, N = 3SE +/- 0.15, N = 3SE +/- 0.80, N = 7SE +/- 0.68, N = 3SE +/- 0.79, N = 12SE +/- 0.31, N = 3SE +/- 0.03, N = 3SE +/- 0.59, N = 12SE +/- 0.65, N = 12SE +/- 0.10, N = 3SE +/- 0.31, N = 3SE +/- 0.23, N = 395.3994.8492.2390.0289.5385.1682.7381.7366.5465.6664.9361.4661.0758.1153.7752.93MAX: 23846.38MAX: 23842.98MAX: 23842.9MAX: 23857.34MAX: 23819.37MAX: 23810.86MAX: 24148.34MAX: 24081.78MAX: 24915.81MAX: 23840.27MAX: 22583.73MAX: 23888.07MAX: 23881.58MAX: 23856.7MAX: 22579.45MAX: 23851.86

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: defconfigEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 9754Xeon Platinum 8592+EPYC 9654EPYC 9684XXeon Max 9468 2PEPYC 9754 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9684X 2P1122334455SE +/- 0.54, N = 4SE +/- 0.50, N = 4SE +/- 0.59, N = 3SE +/- 0.57, N = 3SE +/- 0.55, N = 3SE +/- 0.44, N = 3SE +/- 0.39, N = 4SE +/- 0.43, N = 3SE +/- 0.46, N = 3SE +/- 0.40, N = 4SE +/- 0.28, N = 7SE +/- 0.31, N = 5SE +/- 0.43, N = 3SE +/- 0.33, N = 3SE +/- 0.23, N = 7SE +/- 0.17, N = 1346.3544.2843.4240.5039.4538.5334.5934.5732.8932.1430.6830.5329.9827.6825.8725.87

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Very FastXeon Platinum 8592+Xeon Max 9468Xeon Max 9480Xeon Max 9468 2PEPYC 8534PNEPYC 9754 2PXeon Max 9480 2PXeon Platinum 8490HEPYC 8534PEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9754Xeon Platinum 8490H 2PEPYC 9684X 2PEPYC 9654EPYC 9684X1428425670SE +/- 0.08, N = 3SE +/- 0.12, N = 4SE +/- 0.02, N = 4SE +/- 0.39, N = 4SE +/- 0.06, N = 4SE +/- 0.43, N = 15SE +/- 0.38, N = 9SE +/- 0.01, N = 4SE +/- 0.08, N = 4SE +/- 0.55, N = 5SE +/- 0.16, N = 5SE +/- 0.18, N = 5SE +/- 0.50, N = 5SE +/- 0.12, N = 5SE +/- 0.13, N = 5SE +/- 0.13, N = 535.1039.5640.8244.7544.9246.0046.2549.4250.0453.4453.6654.4654.8855.0562.0262.70

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 128 - Buffer Length: 256 - Filter Length: 57Xeon Max 9468EPYC 8534PNXeon Max 9480EPYC 8534PXeon Platinum 8490HXeon Platinum 8592+Xeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9480 2PEPYC 9654EPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9754EPYC 9684XEPYC 9684X 2P1000M2000M3000M4000M5000MSE +/- 2030052.00, N = 3SE +/- 6398003.16, N = 3SE +/- 3119294.79, N = 3SE +/- 3527510.93, N = 3SE +/- 3116800.35, N = 3SE +/- 5260228.13, N = 3SE +/- 25888414.40, N = 3SE +/- 13655808.69, N = 3SE +/- 14772609.79, N = 3SE +/- 20120967.50, N = 3SE +/- 4098915.85, N = 3SE +/- 51282832.98, N = 4SE +/- 14606429.33, N = 3SE +/- 18691531.06, N = 3SE +/- 6778479.02, N = 3SE +/- 11892621.43, N = 326700333332687066667297240000029828000003181266667350320000035964000003723666667389650000042563000004280933333445052500046265666674714500000475533333347593666671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

ClickHouse

ClickHouse is an open-source, high performance OLAP data management system. This test profile uses ClickHouse's standard benchmark recommendations per https://clickhouse.com/docs/en/operations/performance-test/ / https://github.com/ClickHouse/ClickBench/tree/main/clickhouse with the 100 million rows web analytics dataset. The reported value is the query processing time using the geometric mean of all separate queries performed as an aggregate. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, Third RunEPYC 8534PNEPYC 8534PEPYC 9754 2PXeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9754Xeon Max 9468Xeon Max 9480 2PXeon Max 9480Xeon Platinum 8490HXeon Max 9468 2PXeon Platinum 8490H 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9654EPYC 9684X130260390520650SE +/- 5.23, N = 3SE +/- 1.77, N = 3SE +/- 6.21, N = 5SE +/- 5.39, N = 5SE +/- 3.62, N = 6SE +/- 3.32, N = 3SE +/- 6.03, N = 3SE +/- 3.54, N = 12SE +/- 3.60, N = 3SE +/- 3.45, N = 3SE +/- 3.08, N = 3SE +/- 2.55, N = 3SE +/- 10.00, N = 3SE +/- 2.03, N = 3SE +/- 4.03, N = 3SE +/- 4.34, N = 3335.46338.84427.95438.60458.23458.96470.02471.79476.91481.66501.07501.63525.23526.36559.55595.19MIN: 32.86 / MAX: 5454.55MIN: 35.76 / MAX: 4000MIN: 80.54 / MAX: 5454.55MIN: 58.2 / MAX: 6000MIN: 42.16 / MAX: 6000MIN: 68.65 / MAX: 4000MIN: 32.59 / MAX: 5454.55MIN: 53.43 / MAX: 6000MIN: 36.01 / MAX: 5000MIN: 40.16 / MAX: 5454.55MIN: 54.25 / MAX: 5000MIN: 64.45 / MAX: 6666.67MIN: 81.86 / MAX: 5000MIN: 91.05 / MAX: 6000MIN: 57.03 / MAX: 5000MIN: 62.05 / MAX: 5000

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, Second RunEPYC 8534PEPYC 8534PNEPYC 9754 2PXeon Platinum 8592+ 2PEPYC 9754Xeon Platinum 8592+Xeon Max 9468Xeon Max 9480 2PXeon Max 9480Xeon Platinum 8490HXeon Max 9468 2PXeon Platinum 8490H 2PEPYC 9684X 2PEPYC 9654 2PEPYC 9654EPYC 9684X130260390520650SE +/- 2.99, N = 3SE +/- 5.72, N = 3SE +/- 4.85, N = 5SE +/- 3.10, N = 5SE +/- 9.26, N = 3SE +/- 2.86, N = 6SE +/- 1.94, N = 3SE +/- 3.17, N = 12SE +/- 4.95, N = 3SE +/- 2.93, N = 3SE +/- 0.19, N = 3SE +/- 3.26, N = 3SE +/- 6.78, N = 3SE +/- 8.37, N = 3SE +/- 3.14, N = 3SE +/- 0.42, N = 3337.32337.44420.35443.58450.12455.79465.06467.96474.68486.05499.18507.56528.21529.80562.50597.73MIN: 35.76 / MAX: 4000MIN: 33.44 / MAX: 4615.38MIN: 78.33 / MAX: 4285.71MIN: 62.43 / MAX: 6666.67MIN: 66.15 / MAX: 5454.55MIN: 42.19 / MAX: 6666.67MIN: 32.59 / MAX: 5000MIN: 52.36 / MAX: 6000MIN: 36.43 / MAX: 6000MIN: 40.43 / MAX: 6666.67MIN: 54.2 / MAX: 4615.38MIN: 66.45 / MAX: 5454.55MIN: 92.17 / MAX: 6000MIN: 86.08 / MAX: 5454.55MIN: 57.47 / MAX: 6666.67MIN: 61.41 / MAX: 5454.55

NWChem

NWChem is an open-source high performance computational chemistry package. Per NWChem's documentation, "NWChem aims to provide its users with computational chemistry tools that are scalable both in their ability to treat large scientific computational chemistry problems efficiently, and in their use of available parallel computing resources from high-performance parallel supercomputers to conventional workstation clusters." Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterNWChem 7.0.2Input: C240 BuckyballXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Platinum 8490H 2PXeon Platinum 8592+Xeon Max 9480 2PXeon Platinum 8592+ 2PXeon Max 9468 2PEPYC 9754 2PEPYC 9754EPYC 9654EPYC 9654 2PEPYC 9684XEPYC 9684X 2P50010001500200025002530.52245.22224.62070.32024.71960.91878.31873.11809.71785.01712.61700.01575.51558.41450.51429.91. (F9X) gfortran options: -lnwctask -lccsd -lmcscf -lselci -lmp2 -lmoints -lstepper -ldriver -loptim -lnwdft -lgradients -lcphf -lesp -lddscf -ldangchang -lguess -lhessian -lvib -lnwcutil -lrimp2 -lproperty -lsolvation -lnwints -lprepar -lnwmd -lnwpw -lofpw -lpaw -lpspw -lband -lnwpwlib -lcafe -lspace -lanalyze -lqhop -lpfft -ldplot -ldrdy -lvscf -lqmmm -lqmd -letrans -ltce -lbq -lmm -lcons -lperfm -ldntmc -lccca -ldimqm -lga -larmci -lpeigs -l64to32 -lopenblas -lpthread -lrt -llapack -lnwcblas -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz -lcomex -m64 -ffast-math -std=legacy -fdefault-integer-8 -finline-functions -O2

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Cell Phone Drop TestEPYC 8534PNXeon Max 9468EPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8490H 2PXeon Platinum 8592+EPYC 9754 2PXeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8592+ 2PEPYC 9754EPYC 9654 2PEPYC 9654EPYC 9684X 2PEPYC 9684X816243240SE +/- 0.35, N = 3SE +/- 0.08, N = 3SE +/- 0.18, N = 3SE +/- 0.02, N = 3SE +/- 0.08, N = 3SE +/- 0.10, N = 3SE +/- 0.18, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 3SE +/- 0.32, N = 3SE +/- 0.26, N = 3SE +/- 0.05, N = 3SE +/- 0.26, N = 3SE +/- 0.05, N = 3SE +/- 0.19, N = 335.9533.7533.2331.9231.4928.2227.9027.8627.6327.5126.8823.6923.5322.5322.5220.45

ClickHouse

ClickHouse is an open-source, high performance OLAP data management system. This test profile uses ClickHouse's standard benchmark recommendations per https://clickhouse.com/docs/en/operations/performance-test/ / https://github.com/ClickHouse/ClickBench/tree/main/clickhouse with the 100 million rows web analytics dataset. The reported value is the query processing time using the geometric mean of all separate queries performed as an aggregate. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, First Run / Cold CacheEPYC 8534PNEPYC 8534PEPYC 9754 2PXeon Platinum 8592+ 2PEPYC 9754Xeon Platinum 8592+Xeon Max 9468Xeon Max 9480 2PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8490H 2PXeon Max 9468 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9654EPYC 9684X120240360480600SE +/- 1.88, N = 3SE +/- 1.57, N = 3SE +/- 4.49, N = 5SE +/- 4.12, N = 5SE +/- 2.78, N = 3SE +/- 4.18, N = 6SE +/- 2.85, N = 3SE +/- 3.85, N = 12SE +/- 5.88, N = 3SE +/- 3.41, N = 3SE +/- 4.37, N = 3SE +/- 1.35, N = 3SE +/- 6.05, N = 3SE +/- 4.34, N = 3SE +/- 5.33, N = 3SE +/- 7.88, N = 3322.69323.59411.26423.30433.79437.20446.40460.66461.62465.50485.63491.89504.84513.92536.82566.31MIN: 32.99 / MAX: 4000MIN: 35.69 / MAX: 3750MIN: 77.12 / MAX: 5000MIN: 61.54 / MAX: 6000MIN: 54.1 / MAX: 4285.71MIN: 41.7 / MAX: 5454.55MIN: 32.64 / MAX: 4615.38MIN: 53.62 / MAX: 5454.55MIN: 35.59 / MAX: 5454.55MIN: 40.93 / MAX: 5454.55MIN: 63.63 / MAX: 5000MIN: 51.86 / MAX: 5454.55MIN: 81.97 / MAX: 4615.38MIN: 84.87 / MAX: 5000MIN: 56.6 / MAX: 6666.67MIN: 59.64 / MAX: 5454.55

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: ResNet-50EPYC 8534PNEPYC 8534PXeon Max 9468EPYC 9684X 2PEPYC 9684XXeon Max 9480EPYC 9754Xeon Platinum 8592+EPYC 9754 2PEPYC 9654 2PXeon Max 9480 2PEPYC 9654Xeon Max 9468 2PXeon Platinum 8490HXeon Platinum 8592+ 2PXeon Platinum 8490H 2P4080120160200SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.14, N = 3SE +/- 0.50, N = 3SE +/- 1.39, N = 4SE +/- 0.63, N = 3SE +/- 2.00, N = 12SE +/- 0.58, N = 3SE +/- 0.18, N = 3SE +/- 0.57, N = 3SE +/- 1.04, N = 12SE +/- 0.66, N = 3SE +/- 1.40, N = 3SE +/- 0.03, N = 3SE +/- 0.21, N = 3SE +/- 1.57, N = 392.8298.70119.73120.11123.37123.69125.76129.02131.38133.87135.28138.11141.95143.67150.33161.06

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 1 - Buffer Length: 256 - Filter Length: 512EPYC 9754EPYC 8534PNEPYC 9754 2PEPYC 8534PEPYC 9684X 2PEPYC 9684XEPYC 9654 2PEPYC 9654Xeon Platinum 8490H 2PXeon Platinum 8490HXeon Max 9480 2PXeon Max 9480Xeon Max 9468 2PXeon Max 9468Xeon Platinum 8592+ 2PXeon Platinum 8592+4M8M12M16M20MSE +/- 120841.84, N = 4SE +/- 31571.79, N = 3SE +/- 11050.39, N = 3SE +/- 3464.10, N = 3SE +/- 234395.65, N = 15SE +/- 134727.30, N = 3SE +/- 124115.99, N = 3SE +/- 121304.48, N = 3SE +/- 38670.98, N = 3SE +/- 8685.88, N = 3SE +/- 7571.88, N = 3SE +/- 30179.10, N = 3SE +/- 17457.89, N = 3SE +/- 5238.74, N = 3SE +/- 162096.41, N = 15SE +/- 9492.69, N = 3111745001123633311273667112900001308313313387333133906671339466716843333168823331688400016885333169046671690633318200533187743331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Timed Godot Game Engine Compilation

This test times how long it takes to compile the Godot Game Engine. Godot is a popular, open-source, cross-platform 2D/3D game engine and is built using the SCons build system and targeting the X11 platform. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Godot Game Engine Compilation 4.0Time To CompileEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 9754EPYC 9754 2PXeon Max 9468 2PXeon Platinum 8592+Xeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654Xeon Platinum 8592+ 2PEPYC 9684XEPYC 9654 2PEPYC 9684X 2P4080120160200SE +/- 0.16, N = 3SE +/- 0.32, N = 3SE +/- 0.18, N = 3SE +/- 0.32, N = 3SE +/- 0.76, N = 3SE +/- 0.20, N = 3SE +/- 0.55, N = 3SE +/- 0.56, N = 3SE +/- 0.91, N = 3SE +/- 1.69, N = 3SE +/- 0.56, N = 3SE +/- 0.16, N = 3SE +/- 0.29, N = 3SE +/- 0.12, N = 3SE +/- 0.06, N = 3SE +/- 0.62, N = 3159.62154.01149.38143.81140.23128.28126.11123.88122.72122.70118.57111.69103.38103.35102.5896.16

QMCPACK

QMCPACK is a modern high-performance open-source Quantum Monte Carlo (QMC) simulation code making use of MPI for this benchmark of the H20 example code. QMCPACK is an open-source production level many-body ab initio Quantum Monte Carlo code for computing the electronic structure of atoms, molecules, and solids. QMCPACK is supported by the U.S. Department of Energy. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Execution Time - Seconds, Fewer Is BetterQMCPACK 3.17.1Input: Li2_STO_aeEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480EPYC 9754 2PEPYC 9754Xeon Platinum 8490HEPYC 9654Xeon Max 9480 2PXeon Platinum 8592+EPYC 9654 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9684XEPYC 9684X 2P306090120150SE +/- 0.40, N = 3SE +/- 1.13, N = 3SE +/- 1.20, N = 3SE +/- 1.04, N = 3SE +/- 0.42, N = 3SE +/- 0.63, N = 3SE +/- 0.89, N = 3SE +/- 0.54, N = 3SE +/- 1.11, N = 5SE +/- 1.03, N = 3SE +/- 0.23, N = 3SE +/- 0.88, N = 3SE +/- 1.11, N = 3SE +/- 0.93, N = 3SE +/- 1.06, N = 3SE +/- 0.06, N = 3151.27137.23121.19117.81111.15110.80109.30108.02106.27105.98102.98102.12101.6397.9197.0091.971. (CXX) g++ options: -fopenmp -foffload=disable -finline-limit=1000 -fstrict-aliasing -funroll-all-loops -ffast-math -march=native -O3 -lm -ldl

Timed FFmpeg Compilation

This test times how long it takes to build the FFmpeg multimedia library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed FFmpeg Compilation 7.0Time To CompileEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 9754Xeon Platinum 8592+EPYC 9754 2PXeon Max 9468 2PEPYC 9654Xeon Max 9480 2PEPYC 9684XXeon Platinum 8490H 2PEPYC 9654 2PEPYC 9684X 2PXeon Platinum 8592+ 2P714212835SE +/- 0.04, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.09, N = 3SE +/- 0.04, N = 3SE +/- 0.11, N = 3SE +/- 0.07, N = 3SE +/- 0.30, N = 3SE +/- 0.09, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 328.4227.5426.4624.8424.0623.2221.3621.2220.7920.4920.0119.9919.5118.0917.8417.39

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9754 2PEPYC 8534PNEPYC 8534PEPYC 9684X 2PEPYC 9654 2PEPYC 9754EPYC 9654EPYC 9684X30M60M90M120M150MSE +/- 565297.90, N = 12SE +/- 574699.97, N = 12SE +/- 692701.79, N = 8SE +/- 662950.65, N = 3SE +/- 1037167.36, N = 3SE +/- 308756.34, N = 3SE +/- 761276.42, N = 12SE +/- 746822.47, N = 3SE +/- 666882.79, N = 15SE +/- 412757.79, N = 3SE +/- 13050.78, N = 3SE +/- 353097.82, N = 3SE +/- 1152766.29, N = 3SE +/- 68559.13, N = 3SE +/- 255887.16, N = 3SE +/- 371659.60, N = 37750124878057728801696208245566483475645877972068923185092709751103630483105138551105432862106778379108821580116685326125562060126161028

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Medium Mesh Size - Mesh TimeEPYC 9754 2PEPYC 8534PNEPYC 9754EPYC 8534PXeon Platinum 8490H 2PXeon Platinum 8490HXeon Max 9480 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8592+EPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9684X 2PXeon Max 9468 2PEPYC 9684X4080120160200192.42188.37167.41163.70162.03160.73159.31153.12150.67148.40140.48139.62136.40135.52134.78118.541. (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

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: Noise-GaussianEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9468 2PEPYC 9754Xeon Max 9480Xeon Platinum 8490HEPYC 9654EPYC 9754 2PXeon Max 9480 2PEPYC 9684XXeon Platinum 8592+EPYC 9654 2PEPYC 9684X 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2P4080120160200SE +/- 0.00, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.58, N = 3SE +/- 0.58, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.58, N = 3SE +/- 0.58, N = 3SE +/- 0.33, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.67, N = 3SE +/- 0.67, N = 3SE +/- 1.15, N = 3SE +/- 0.88, N = 31201231501501511581581601641641651711781791821941. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: AlexNetEPYC 8534PNEPYC 9754 2PEPYC 8534PEPYC 9654 2PEPYC 9684X 2PXeon Platinum 8592+Xeon Max 9468Xeon Max 9480Xeon Platinum 8592+ 2PEPYC 9684XEPYC 9754Xeon Max 9468 2PXeon Max 9480 2PEPYC 9654Xeon Platinum 8490HXeon Platinum 8490H 2P2004006008001000SE +/- 0.16, N = 4SE +/- 7.43, N = 15SE +/- 0.45, N = 4SE +/- 8.87, N = 4SE +/- 7.66, N = 5SE +/- 6.22, N = 5SE +/- 1.95, N = 5SE +/- 2.31, N = 5SE +/- 2.22, N = 5SE +/- 1.37, N = 5SE +/- 2.99, N = 5SE +/- 6.39, N = 13SE +/- 7.35, N = 5SE +/- 1.31, N = 5SE +/- 0.51, N = 5SE +/- 7.98, N = 5617.03658.97673.42727.77738.31819.89841.68852.36858.84868.54924.80932.59940.40942.19969.62996.47

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 1 - Buffer Length: 256 - Filter Length: 57EPYC 8534PNEPYC 8534PEPYC 9754 2PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9654EPYC 9684XXeon Max 9480 2PXeon Platinum 8490H 2PXeon Max 9468 2PXeon Platinum 8490HXeon Max 9480Xeon Max 9468Xeon Platinum 8592+ 2PXeon Platinum 8592+16M32M48M64M80MSE +/- 393224.42, N = 3SE +/- 4807.40, N = 3SE +/- 8089.77, N = 3SE +/- 7937.25, N = 3SE +/- 486361.09, N = 3SE +/- 34724.31, N = 3SE +/- 12719.19, N = 3SE +/- 7423.69, N = 3SE +/- 822374.71, N = 4SE +/- 3844.19, N = 3SE +/- 4163.33, N = 3SE +/- 666.67, N = 3SE +/- 5364.49, N = 3SE +/- 4910.31, N = 3SE +/- 68119.99, N = 3SE +/- 599433.09, N = 9454863334621066746229667462700005495466755415667554253335545533366266000669913336706400067074667671276676713033370499000731406671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 8 - Input: Bosphorus 4KXeon Max 9468 2PXeon Max 9480 2PXeon Max 9468EPYC 8534PNXeon Platinum 8490H 2PXeon Max 9480EPYC 8534PXeon Platinum 8592+ 2PXeon Platinum 8490HEPYC 9754 2PEPYC 9754Xeon Platinum 8592+EPYC 9684XEPYC 9654 2PEPYC 9684X 2PEPYC 965420406080100SE +/- 0.58, N = 5SE +/- 0.44, N = 15SE +/- 0.26, N = 4SE +/- 0.67, N = 4SE +/- 0.46, N = 4SE +/- 0.29, N = 4SE +/- 0.11, N = 4SE +/- 0.79, N = 5SE +/- 0.20, N = 5SE +/- 0.31, N = 4SE +/- 0.38, N = 5SE +/- 0.22, N = 5SE +/- 0.42, N = 5SE +/- 0.40, N = 5SE +/- 0.63, N = 5SE +/- 0.56, N = 554.8255.9862.7564.7965.2265.3668.1970.4370.9177.6078.7479.4486.3086.3487.8287.991. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Timed Gem5 Compilation

This test times how long it takes to compile Gem5. Gem5 is a simulator for computer system architecture research. Gem5 is widely used for computer architecture research within the industry, academia, and more. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Gem5 Compilation 23.0.1Time To CompileEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480EPYC 9754Xeon Platinum 8490HEPYC 9754 2PXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9684XEPYC 9654 2PEPYC 9684X 2PXeon Platinum 8592+ 2P50100150200250SE +/- 0.97, N = 3SE +/- 1.99, N = 9SE +/- 2.59, N = 4SE +/- 2.11, N = 6SE +/- 1.88, N = 3SE +/- 0.92, N = 3SE +/- 1.73, N = 3SE +/- 2.17, N = 3SE +/- 2.24, N = 4SE +/- 1.72, N = 3SE +/- 1.01, N = 3SE +/- 0.58, N = 3SE +/- 2.15, N = 3SE +/- 0.72, N = 3SE +/- 1.16, N = 3SE +/- 0.89, N = 3251.61242.93220.57215.50212.55210.93192.98187.51185.30184.30181.63178.28176.88164.90161.94157.36

Timed ImageMagick Compilation

This test times how long it takes to build ImageMagick. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed ImageMagick Compilation 6.9.0Time To CompileEPYC 8534PNEPYC 8534PEPYC 9754Xeon Max 9468EPYC 9754 2PXeon Max 9480Xeon Platinum 8490HXeon Max 9468 2PXeon Max 9480 2PEPYC 9654Xeon Platinum 8490H 2PEPYC 9654 2PXeon Platinum 8592+EPYC 9684XEPYC 9684X 2PXeon Platinum 8592+ 2P48121620SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.08, N = 4SE +/- 0.15, N = 4SE +/- 0.02, N = 4SE +/- 0.06, N = 4SE +/- 0.10, N = 4SE +/- 0.11, N = 4SE +/- 0.07, N = 4SE +/- 0.09, N = 4SE +/- 0.13, N = 4SE +/- 0.04, N = 4SE +/- 0.10, N = 4SE +/- 0.06, N = 4SE +/- 0.04, N = 4SE +/- 0.03, N = 418.1817.9216.3916.0815.8115.3814.7214.4014.0913.4713.2612.7512.6012.2011.7211.49

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9754 2PEPYC 8534PEPYC 8534PNEPYC 9684X 2PEPYC 9654 2PEPYC 9754EPYC 9654EPYC 9684X70140210280350SE +/- 2.59, N = 12SE +/- 3.74, N = 12SE +/- 3.79, N = 8SE +/- 5.15, N = 3SE +/- 5.07, N = 3SE +/- 2.20, N = 3SE +/- 3.40, N = 12SE +/- 3.41, N = 3SE +/- 2.16, N = 15SE +/- 0.38, N = 3SE +/- 0.82, N = 3SE +/- 1.87, N = 3SE +/- 3.49, N = 3SE +/- 1.41, N = 3SE +/- 1.96, N = 3SE +/- 1.71, N = 3328.41325.97317.36314.44309.03294.33286.78280.26253.13251.47251.05246.08240.43227.40211.85209.68MAX: 29309.84MAX: 30042.47MAX: 32341.07MAX: 27244.84MAX: 26720.06MAX: 27740.38MAX: 28496.67MAX: 27450.82MAX: 27854.54MAX: 26615.75MAX: 26590.2MAX: 27665.32MAX: 27651.54MAX: 26860.34MAX: 26989.76MAX: 26606.61

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 64 - Buffer Length: 256 - Filter Length: 512EPYC 8534PNXeon Max 9468EPYC 8534PXeon Max 9480EPYC 9754 2PEPYC 9754Xeon Max 9468 2PXeon Platinum 8490HXeon Max 9480 2PXeon Platinum 8592+Xeon Platinum 8490H 2PEPYC 9654EPYC 9654 2PEPYC 9684X 2PEPYC 9684XXeon Platinum 8592+ 2P200M400M600M800M1000MSE +/- 2292977.20, N = 3SE +/- 1611276.24, N = 3SE +/- 822739.33, N = 3SE +/- 1289603.73, N = 3SE +/- 958772.36, N = 3SE +/- 673003.22, N = 3SE +/- 4432294.36, N = 3SE +/- 940265.92, N = 3SE +/- 6141471.96, N = 3SE +/- 1515552.70, N = 3SE +/- 3717554.45, N = 3SE +/- 1678355.68, N = 3SE +/- 2261241.06, N = 3SE +/- 3039034.93, N = 3SE +/- 752558.60, N = 3SE +/- 7562646.32, N = 85772633336323966676338600006890033337093933337096500007325600007444100007571766677837700007994966678113733338245833338493400008505866678837012501. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Disney MaterialEPYC 8534PNXeon Max 9468EPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+EPYC 9754Xeon Platinum 8592+ 2PEPYC 9654EPYC 9684X142842567061.3658.3856.4256.4053.4951.2549.1448.5648.4345.0843.7043.4540.92

Scene: Disney Material

EPYC 9654 2P: The test run did not produce a result.

EPYC 9684X 2P: The test run did not produce a result.

EPYC 9754 2P: The test run did not produce a result.

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19, Long Mode - Compression SpeedEPYC 9754 2PEPYC 9754EPYC 8534PNEPYC 8534PXeon Max 9480 2PXeon Max 9468 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8490HXeon Platinum 8490H 2PEPYC 9654 2PEPYC 9654EPYC 9684X 2PEPYC 9684XXeon Platinum 8592+ 2PXeon Platinum 8592+3691215SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.06, N = 15SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.05, N = 36.646.686.746.767.467.627.727.767.948.168.308.388.428.459.679.801. (CC) gcc options: -O3 -pthread -lz -llzma

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Face Detection Retail FP16-INT8 - Device: CPUEPYC 8534PNEPYC 8534PEPYC 9754 2PXeon Platinum 8490HEPYC 9754Xeon Max 9480 2PEPYC 9654 2PEPYC 9654Xeon Max 9480Xeon Platinum 8490H 2PXeon Max 9468 2PEPYC 9684XEPYC 9684X 2PXeon Max 9468Xeon Platinum 8592+Xeon Platinum 8592+ 2P246810SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 37.606.776.606.436.256.025.985.855.775.725.555.415.345.275.195.18MIN: 3.84 / MAX: 21.81MIN: 3.54 / MAX: 18.38MIN: 5.84 / MAX: 35.85MIN: 3.84 / MAX: 15.83MIN: 3.07 / MAX: 31.02MIN: 5.37 / MAX: 27.77MIN: 4.9 / MAX: 26.69MIN: 2.67 / MAX: 22.66MIN: 5.26 / MAX: 13.86MIN: 5.1 / MAX: 17.93MIN: 5.12 / MAX: 24.99MIN: 2.65 / MAX: 24.3MIN: 4.88 / MAX: 26.62MIN: 4.98 / MAX: 13.16MIN: 4.72 / MAX: 11.02MIN: 4.74 / MAX: 26.351. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 12 - Input: Bosphorus 4KXeon Max 9468 2PXeon Max 9480 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8490H 2PEPYC 8534PNEPYC 9684X 2PXeon Platinum 8490HEPYC 8534PXeon Platinum 8592+ 2PEPYC 9684XEPYC 9754 2PEPYC 9754EPYC 9654 2PXeon Platinum 8592+EPYC 96544080120160200SE +/- 0.83, N = 5SE +/- 0.97, N = 6SE +/- 0.61, N = 7SE +/- 0.40, N = 7SE +/- 0.25, N = 7SE +/- 0.78, N = 6SE +/- 0.88, N = 6SE +/- 0.41, N = 7SE +/- 0.71, N = 6SE +/- 0.42, N = 7SE +/- 0.43, N = 7SE +/- 1.41, N = 6SE +/- 1.32, N = 6SE +/- 1.88, N = 15SE +/- 0.20, N = 7SE +/- 0.80, N = 7122.09127.55146.10151.72156.52157.36163.80164.25165.09165.50165.75168.78169.23171.22177.86178.521. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19 - Decompression SpeedXeon Platinum 8490H 2PXeon Platinum 8490HXeon Max 9480 2PXeon Max 9468 2PEPYC 8534PXeon Max 9468Xeon Max 9480EPYC 9754 2PEPYC 9754EPYC 8534PNXeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9654 2PEPYC 9684XEPYC 9654EPYC 9684X 2P30060090012001500SE +/- 4.96, N = 3SE +/- 4.45, N = 3SE +/- 0.65, N = 15SE +/- 0.66, N = 15SE +/- 21.23, N = 3SE +/- 2.12, N = 3SE +/- 1.08, N = 15SE +/- 1.83, N = 3SE +/- 0.83, N = 3SE +/- 1.58, N = 3SE +/- 1.88, N = 3SE +/- 3.01, N = 4SE +/- 2.20, N = 3SE +/- 7.31, N = 3SE +/- 0.46, N = 3SE +/- 1.73, N = 3963.61032.01039.21040.21118.81126.41127.51139.01139.91144.81165.81221.81382.11385.31388.91392.51. (CC) gcc options: -O3 -pthread -lz -llzma

x265

This is a simple test of the x265 encoder run on the CPU with 1080p and 4K options for H.265 video encode performance with x265. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.6Video Input: Bosphorus 4KXeon Max 9468 2PEPYC 8534PNXeon Max 9480 2PEPYC 9754 2PEPYC 8534PEPYC 9754EPYC 9654 2PXeon Max 9468Xeon Max 9480EPYC 9654EPYC 9684X 2PEPYC 9684XXeon Platinum 8490H 2PXeon Platinum 8490HXeon Platinum 8592+Xeon Platinum 8592+ 2P918273645SE +/- 0.12, N = 3SE +/- 0.06, N = 3SE +/- 0.10, N = 3SE +/- 0.06, N = 3SE +/- 0.06, N = 3SE +/- 0.02, N = 3SE +/- 0.09, N = 3SE +/- 0.08, N = 3SE +/- 0.39, N = 3SE +/- 0.08, N = 3SE +/- 0.10, N = 3SE +/- 0.11, N = 3SE +/- 0.20, N = 4SE +/- 0.11, N = 4SE +/- 0.06, N = 4SE +/- 0.09, N = 427.2627.7427.9028.1128.1328.1628.8129.0531.4031.7231.9532.1237.0437.8638.9339.311. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 4 - Input: Bosphorus 4KXeon Max 9468 2PXeon Max 9480 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8490H 2PEPYC 8534PNEPYC 8534PXeon Platinum 8490HEPYC 9754 2PEPYC 9754Xeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9654 2PEPYC 9654EPYC 9684X 2PEPYC 9684X246810SE +/- 0.015, N = 3SE +/- 0.042, N = 3SE +/- 0.027, N = 3SE +/- 0.047, N = 3SE +/- 0.049, N = 3SE +/- 0.015, N = 3SE +/- 0.007, N = 3SE +/- 0.030, N = 3SE +/- 0.004, N = 3SE +/- 0.024, N = 3SE +/- 0.044, N = 3SE +/- 0.005, N = 3SE +/- 0.046, N = 3SE +/- 0.019, N = 3SE +/- 0.040, N = 3SE +/- 0.021, N = 35.9916.0526.4906.5726.6986.7276.9397.0617.1537.2837.3607.8768.3498.4228.5138.5951. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.11Video Input: Bosphorus 4K - Video Preset: FasterXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480EPYC 9754Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654EPYC 9684X48121620SE +/- 0.11, N = 3SE +/- 0.13, N = 3SE +/- 0.16, N = 3SE +/- 0.12, N = 3SE +/- 0.01, N = 3SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.13, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.08, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 311.9011.9712.5413.2613.6113.9714.1014.3614.7915.0615.9716.7317.021. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

Video Input: Bosphorus 4K - Video Preset: Faster

EPYC 9654 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1

EPYC 9684X 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1

EPYC 9754 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Small Mesh Size - Mesh TimeEPYC 8534PNXeon Platinum 8490H 2PEPYC 8534PXeon Max 9468 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8592+Xeon Platinum 8490HEPYC 9754EPYC 9754 2PXeon Platinum 8592+ 2PXeon Max 9480 2PEPYC 9654EPYC 9654 2PEPYC 9684X 2PEPYC 9684X81624324036.8235.6033.6732.4832.0031.7831.7031.3830.3829.4729.3328.8428.0327.7426.8225.931. (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

PyBench

This test profile reports the total time of the different average timed test results from PyBench. PyBench reports average test times for different functions such as BuiltinFunctionCalls and NestedForLoops, with this total result providing a rough estimate as to Python's average performance on a given system. This test profile runs PyBench each time for 20 rounds. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyBench 2018-02-16Total For Average Test TimesEPYC 8534PEPYC 9754EPYC 8534PNEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9684XEPYC 9654Xeon Platinum 8490H 2PXeon Platinum 8490HXeon Max 9480 2PXeon Max 9468 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8592+ 2PXeon Platinum 8592+2004006008001000SE +/- 1.45, N = 3SE +/- 0.58, N = 3SE +/- 2.40, N = 3SE +/- 1.33, N = 3SE +/- 1.86, N = 3SE +/- 1.20, N = 3SE +/- 2.08, N = 3SE +/- 2.19, N = 3SE +/- 1.15, N = 3SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 0.67, N = 3SE +/- 0.33, N = 3SE +/- 0.58, N = 3SE +/- 0.48, N = 4SE +/- 0.41, N = 4976974974972822818818818767765763763763762692691

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 64 - Buffer Length: 256 - Filter Length: 57EPYC 8534PNXeon Max 9468EPYC 8534PXeon Max 9480EPYC 9754 2PEPYC 9754Xeon Platinum 8592+Xeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490HXeon Platinum 8592+ 2PXeon Platinum 8490H 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9654EPYC 9684X600M1200M1800M2400M3000MSE +/- 7776531.65, N = 3SE +/- 2629533.12, N = 3SE +/- 4914717.13, N = 3SE +/- 5493127.02, N = 3SE +/- 17457503.96, N = 3SE +/- 8567444.84, N = 3SE +/- 11426868.92, N = 3SE +/- 27382304.50, N = 5SE +/- 28812627.44, N = 4SE +/- 7076957.92, N = 3SE +/- 32405829.38, N = 3SE +/- 28500058.48, N = 3SE +/- 10590929.03, N = 3SE +/- 8732315.46, N = 3SE +/- 3786085.64, N = 3SE +/- 8476044.68, N = 320263333332138766667218683333323089666672374766667244493333324755000002491340000249405000025392000002620066667265550000027302333332772800000285983333328620000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 64 - Buffer Length: 256 - Filter Length: 32Xeon Max 9468Xeon Max 9480EPYC 8534PNEPYC 8534PXeon Platinum 8490HEPYC 9754 2PEPYC 9754Xeon Platinum 8592+Xeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9480 2PXeon Platinum 8592+ 2PEPYC 9684X 2PEPYC 9654 2PEPYC 9654EPYC 9684X500M1000M1500M2000M2500MSE +/- 688799.28, N = 3SE +/- 692820.32, N = 3SE +/- 3256958.36, N = 3SE +/- 3090487.20, N = 3SE +/- 3167192.94, N = 3SE +/- 1462114.15, N = 3SE +/- 1550627.11, N = 3SE +/- 27191440.65, N = 3SE +/- 2206304.10, N = 3SE +/- 3300505.01, N = 3SE +/- 3017909.50, N = 3SE +/- 17236814.61, N = 3SE +/- 2510644.01, N = 3SE +/- 1567730.14, N = 3SE +/- 2810891.52, N = 3SE +/- 3000185.18, N = 316402666671679700000177686666718713333331890566667189496666718987666671927766667196323333320250000002090466667215463333322793000002282066667228276666722874666671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

QuantLib

QuantLib is an open-source library/framework around quantitative finance for modeling, trading and risk management scenarios. QuantLib is written in C++ with Boost and its built-in benchmark used reports the QuantLib Benchmark Index benchmark score. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMFLOPS, More Is BetterQuantLib 1.32Configuration: Single-ThreadedEPYC 9754 2PEPYC 8534PNEPYC 8534PEPYC 9754EPYC 9654 2PEPYC 9654EPYC 9684XEPYC 9684X 2PXeon Max 9468 2PXeon Platinum 8490HXeon Platinum 8490H 2PXeon Max 9480 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8592+ 2PXeon Platinum 8592+8001600240032004000SE +/- 19.30, N = 3SE +/- 14.40, N = 3SE +/- 18.64, N = 3SE +/- 14.83, N = 3SE +/- 19.83, N = 3SE +/- 24.27, N = 3SE +/- 22.39, N = 3SE +/- 1.14, N = 3SE +/- 28.11, N = 3SE +/- 28.42, N = 3SE +/- 25.40, N = 3SE +/- 28.57, N = 3SE +/- 22.17, N = 3SE +/- 25.35, N = 3SE +/- 30.36, N = 12SE +/- 33.58, N = 32590.82593.02593.32596.13098.53105.93114.23125.43207.03209.63210.13222.83225.73229.13586.33600.51. (CXX) g++ options: -O3 -march=native -fPIE -pie

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: ParallelEPYC 9684X 2PXeon Max 9480 2PXeon Max 9468 2PEPYC 9654 2PEPYC 9754 2PEPYC 9754EPYC 8534PNEPYC 8534PXeon Max 9480Xeon Max 9468EPYC 9684XEPYC 9654Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PXeon Platinum 8490HXeon Platinum 8592+816243240SE +/- 0.08, N = 3SE +/- 0.19, N = 3SE +/- 0.34, N = 3SE +/- 0.05, N = 3SE +/- 0.24, N = 7SE +/- 0.33, N = 4SE +/- 0.25, N = 3SE +/- 0.14, N = 3SE +/- 0.07, N = 3SE +/- 0.10, N = 3SE +/- 0.19, N = 3SE +/- 0.36, N = 3SE +/- 0.06, N = 3SE +/- 0.00, N = 3SE +/- 0.07, N = 3SE +/- 0.01, N = 324.0524.6824.7325.6326.3327.4527.8727.9428.4828.5228.7329.8430.0431.7132.3433.381. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 6Xeon Max 9468 2PXeon Max 9480 2PXeon Max 9468EPYC 8534PNXeon Max 9480EPYC 8534PEPYC 9754 2PEPYC 9754Xeon Platinum 8490HXeon Platinum 8490H 2PXeon Platinum 8592+EPYC 9684X 2PEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9684X0.84961.69922.54883.39844.248SE +/- 0.029, N = 8SE +/- 0.022, N = 9SE +/- 0.017, N = 9SE +/- 0.011, N = 9SE +/- 0.011, N = 9SE +/- 0.016, N = 9SE +/- 0.026, N = 9SE +/- 0.002, N = 9SE +/- 0.022, N = 9SE +/- 0.008, N = 9SE +/- 0.016, N = 9SE +/- 0.005, N = 9SE +/- 0.008, N = 10SE +/- 0.009, N = 10SE +/- 0.011, N = 10SE +/- 0.005, N = 103.7763.7383.7303.6233.5503.4843.3413.2513.1693.0442.9712.8902.8102.8052.7562.7231. (CXX) g++ options: -O3 -fPIC -lm

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PDSCH Processor Benchmark, Throughput ThreadEPYC 8534PEPYC 8534PNEPYC 9754 2PEPYC 9754Xeon Max 9468 2PXeon Max 9480 2PEPYC 9684X 2PEPYC 9654 2PXeon Platinum 8490H 2PXeon Platinum 8490HEPYC 9684XEPYC 9654Xeon Max 9468Xeon Max 9480Xeon Platinum 8592+Xeon Platinum 8592+ 2P160320480640800SE +/- 3.14, N = 7SE +/- 1.24, N = 7SE +/- 4.48, N = 7SE +/- 0.80, N = 7SE +/- 3.86, N = 8SE +/- 5.02, N = 8SE +/- 5.52, N = 8SE +/- 3.72, N = 8SE +/- 2.57, N = 8SE +/- 1.17, N = 8SE +/- 3.13, N = 8SE +/- 1.71, N = 8SE +/- 1.64, N = 8SE +/- 1.27, N = 8SE +/- 1.34, N = 8SE +/- 1.89, N = 8557.9563.6576.8596.2651.9663.8664.7673.4675.5680.2680.7684.7702.6705.5742.9755.51. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -mavx512f -mavx512cd -mavx512bw -mavx512dq -fno-trapping-math -fno-math-errno -ldl

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19 - Compression SpeedXeon Max 9480 2PEPYC 9754 2PEPYC 8534PNEPYC 9754Xeon Max 9468 2PEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8490H 2PEPYC 9684X 2PEPYC 9654 2PEPYC 9654EPYC 9684XXeon Platinum 8592+ 2PXeon Platinum 8592+48121620SE +/- 0.13, N = 15SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.06, N = 3SE +/- 0.09, N = 15SE +/- 0.00, N = 3SE +/- 0.15, N = 3SE +/- 0.11, N = 15SE +/- 0.06, N = 3SE +/- 0.12, N = 3SE +/- 0.22, N = 3SE +/- 0.09, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.21, N = 413.313.713.813.813.813.914.414.514.615.216.817.017.217.217.818.01. (CC) gcc options: -O3 -pthread -lz -llzma

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 1 - Buffer Length: 256 - Filter Length: 32EPYC 9754 2PEPYC 8534PNEPYC 8534PEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9684XEPYC 9654Xeon Platinum 8490H 2PXeon Max 9468 2PXeon Platinum 8490HXeon Max 9480 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8592+ 2PXeon Platinum 8592+9M18M27M36M45MSE +/- 2185.81, N = 3SE +/- 881.92, N = 3SE +/- 2333.33, N = 3SE +/- 6984.11, N = 3SE +/- 13503.09, N = 3SE +/- 19139.84, N = 3SE +/- 7023.77, N = 3SE +/- 7753.14, N = 3SE +/- 1154.70, N = 3SE +/- 1201.85, N = 3SE +/- 1527.53, N = 3SE +/- 1452.97, N = 3SE +/- 1666.67, N = 3SE +/- 3000.00, N = 3SE +/- 419600.12, N = 5SE +/- 417042.80, N = 5298933332989933329899667299073333603300036040000360660003606833336461000364976673650600036521333365376673653900038950600402440001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Zstd Compression

This test measures the time needed to compress/decompress a sample file (silesia.tar) using Zstd (Zstandard) compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.4Compression Level: 19, Long Mode - Decompression SpeedXeon Platinum 8490H 2PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490HEPYC 8534PNEPYC 8534PEPYC 9754EPYC 9754 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9684X 2PEPYC 9654 2PEPYC 9684XEPYC 965430060090012001500SE +/- 3.69, N = 3SE +/- 1.00, N = 3SE +/- 1.78, N = 15SE +/- 0.43, N = 3SE +/- 0.92, N = 3SE +/- 2.09, N = 3SE +/- 1.95, N = 3SE +/- 0.19, N = 3SE +/- 0.56, N = 3SE +/- 1.45, N = 3SE +/- 1.60, N = 3SE +/- 0.84, N = 3SE +/- 2.93, N = 3SE +/- 1.87, N = 3SE +/- 1.99, N = 3SE +/- 1.17, N = 3992.11041.81046.01069.11096.51101.41102.21104.21131.11133.31189.71242.61321.41324.61324.91328.51. (CC) gcc options: -O3 -pthread -lz -llzma

Timed Mesa Compilation

This test profile times how long it takes to compile Mesa with Meson/Ninja. For minimizing build dependencies and avoid versioning conflicts, test this is just the core Mesa build without LLVM or the extra Gallium3D/Mesa drivers enabled. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Mesa Compilation 24.0Time To CompileEPYC 8534PNEPYC 8534PXeon Max 9468EPYC 9754Xeon Max 9480EPYC 9754 2PXeon Platinum 8490HXeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9480 2PXeon Platinum 8592+EPYC 9654EPYC 9684XEPYC 9654 2PEPYC 9684X 2PXeon Platinum 8592+ 2P510152025SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.08, N = 3SE +/- 0.12, N = 3SE +/- 0.07, N = 3SE +/- 0.07, N = 3SE +/- 0.12, N = 3SE +/- 0.17, N = 3SE +/- 0.07, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 4SE +/- 0.08, N = 4SE +/- 0.05, N = 421.5321.0420.1819.9119.6519.5619.2318.4718.2918.1417.2616.9616.6716.4116.4016.18

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 6, LosslessEPYC 8534PNEPYC 8534PXeon Max 9468 2PXeon Max 9480 2PEPYC 9754 2PEPYC 9754Xeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8490H 2PEPYC 9684X 2PEPYC 9654 2PEPYC 9654Xeon Platinum 8592+EPYC 9684XXeon Platinum 8592+ 2P246810SE +/- 0.011, N = 6SE +/- 0.009, N = 6SE +/- 0.027, N = 6SE +/- 0.035, N = 6SE +/- 0.025, N = 6SE +/- 0.007, N = 6SE +/- 0.045, N = 6SE +/- 0.033, N = 6SE +/- 0.013, N = 7SE +/- 0.022, N = 7SE +/- 0.015, N = 7SE +/- 0.020, N = 7SE +/- 0.010, N = 7SE +/- 0.025, N = 7SE +/- 0.014, N = 7SE +/- 0.031, N = 77.3017.1827.1217.0386.9866.8536.8366.6056.2786.1725.9505.8465.7995.7845.7575.6131. (CXX) g++ options: -O3 -fPIC -lm

Numpy Benchmark

This is a test to obtain the general Numpy performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterNumpy BenchmarkEPYC 8534PNEPYC 8534PEPYC 9754 2PEPYC 9754Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490HXeon Platinum 8490H 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9654 2PEPYC 9654EPYC 9684X 2PEPYC 9684X120240360480600SE +/- 1.34, N = 3SE +/- 0.72, N = 3SE +/- 0.25, N = 3SE +/- 0.96, N = 3SE +/- 1.13, N = 3SE +/- 1.30, N = 3SE +/- 0.33, N = 3SE +/- 0.26, N = 3SE +/- 0.60, N = 3SE +/- 0.79, N = 3SE +/- 1.42, N = 3SE +/- 1.84, N = 3SE +/- 0.38, N = 3SE +/- 2.28, N = 3SE +/- 0.60, N = 3SE +/- 1.85, N = 3418.89421.99422.76425.46435.21439.84455.57456.18461.61464.72519.89520.22529.03529.80540.96542.15

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 0EPYC 8534PNXeon Max 9468EPYC 8534PEPYC 9754 2PEPYC 9754Xeon Max 9480Xeon Platinum 8490HXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+EPYC 9684X 2PEPYC 9654 2PEPYC 9654EPYC 9684XXeon Platinum 8592+ 2P20406080100SE +/- 0.18, N = 3SE +/- 0.15, N = 3SE +/- 0.19, N = 3SE +/- 0.06, N = 3SE +/- 0.13, N = 3SE +/- 0.22, N = 3SE +/- 0.32, N = 3SE +/- 0.29, N = 3SE +/- 0.32, N = 3SE +/- 0.05, N = 3SE +/- 0.08, N = 3SE +/- 0.01, N = 3SE +/- 0.11, N = 3SE +/- 0.01, N = 3SE +/- 0.12, N = 3SE +/- 0.03, N = 378.2476.9275.8674.6973.2773.0669.9369.6568.4467.3864.4863.4062.8961.9961.6360.531. (CXX) g++ options: -O3 -fPIC -lm

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 10, LosslessEPYC 9754 2PEPYC 8534PEPYC 8534PNEPYC 9754Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9684X 2PXeon Platinum 8490HXeon Max 9468Xeon Max 9480EPYC 9654 2PEPYC 9654EPYC 9684XXeon Platinum 8592+ 2PXeon Platinum 8592+1.29152.5833.87455.1666.4575SE +/- 0.006, N = 7SE +/- 0.006, N = 7SE +/- 0.010, N = 7SE +/- 0.004, N = 7SE +/- 0.019, N = 7SE +/- 0.028, N = 7SE +/- 0.014, N = 7SE +/- 0.003, N = 7SE +/- 0.006, N = 7SE +/- 0.005, N = 7SE +/- 0.015, N = 7SE +/- 0.006, N = 8SE +/- 0.006, N = 8SE +/- 0.002, N = 8SE +/- 0.015, N = 8SE +/- 0.008, N = 85.7405.7355.7345.6625.3175.3115.0744.9434.9234.9164.8674.8294.7924.7704.5794.4561. (CXX) g++ options: -O3 -fPIC -lm

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 2EPYC 8534PNEPYC 9754 2PEPYC 8534PXeon Max 9468EPYC 9754Xeon Max 9480Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490HXeon Platinum 8490H 2PEPYC 9654 2PEPYC 9684X 2PXeon Platinum 8592+EPYC 9654EPYC 9684XXeon Platinum 8592+ 2P1020304050SE +/- 0.10, N = 3SE +/- 0.13, N = 3SE +/- 0.03, N = 3SE +/- 0.21, N = 3SE +/- 0.06, N = 3SE +/- 0.31, N = 3SE +/- 0.22, N = 3SE +/- 0.06, N = 3SE +/- 0.04, N = 3SE +/- 0.16, N = 3SE +/- 0.43, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.09, N = 342.6141.9241.6641.1340.3539.8838.6538.1437.7537.3135.4035.3234.7934.1534.0433.651. (CXX) g++ options: -O3 -fPIC -lm

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.11Video Input: Bosphorus 4K - Video Preset: FastEPYC 8534PNXeon Max 9468 2PXeon Max 9480 2PEPYC 8534PXeon Max 9468EPYC 9754Xeon Platinum 8490H 2PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9654EPYC 9684X246810SE +/- 0.041, N = 3SE +/- 0.081, N = 3SE +/- 0.043, N = 15SE +/- 0.038, N = 3SE +/- 0.010, N = 3SE +/- 0.045, N = 3SE +/- 0.047, N = 3SE +/- 0.033, N = 3SE +/- 0.022, N = 3SE +/- 0.038, N = 3SE +/- 0.023, N = 3SE +/- 0.048, N = 3SE +/- 0.057, N = 36.0006.0236.0776.1186.3246.4656.4806.5016.7196.9607.3927.4317.5471. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

Video Input: Bosphorus 4K - Video Preset: Fast

EPYC 9654 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1

EPYC 9684X 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1

EPYC 9754 2P: The test quit with a non-zero exit status. E: Parameter Check Error: Number of threads out of range (-1

WebP Image Encode

This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.2.4Encode Settings: Quality 100, Lossless, Highest CompressionEPYC 8534PEPYC 8534PNEPYC 9754EPYC 9754 2PXeon Platinum 8490H 2PXeon Platinum 8490HXeon Max 9468 2PXeon Max 9480 2PXeon Max 9468Xeon Max 9480EPYC 9684X 2PEPYC 9654EPYC 9654 2PEPYC 9684XXeon Platinum 8592+ 2PXeon Platinum 8592+0.1350.270.4050.540.675SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.480.480.480.480.510.520.540.540.550.550.560.570.570.570.590.601. (CC) gcc options: -fvisibility=hidden -O2 -lm

SecureMark

SecureMark is an objective, standardized benchmarking framework for measuring the efficiency of cryptographic processing solutions developed by EEMBC. SecureMark-TLS is benchmarking Transport Layer Security performance with a focus on IoT/edge computing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmarks, More Is BetterSecureMark 1.0.4Benchmark: SecureMark-TLSEPYC 8534PEPYC 8534PNEPYC 9754EPYC 9754 2PXeon Platinum 8490H 2PXeon Platinum 8490HXeon Max 9480Xeon Max 9480 2PXeon Max 9468 2PXeon Max 9468EPYC 9654EPYC 9654 2PEPYC 9684XEPYC 9684X 2PXeon Platinum 8592+ 2PXeon Platinum 8592+70K140K210K280K350KSE +/- 372.96, N = 3SE +/- 542.88, N = 3SE +/- 457.42, N = 3SE +/- 553.62, N = 3SE +/- 330.95, N = 3SE +/- 625.56, N = 3SE +/- 670.43, N = 3SE +/- 196.24, N = 3SE +/- 384.08, N = 3SE +/- 256.85, N = 3SE +/- 361.20, N = 3SE +/- 389.81, N = 3SE +/- 415.71, N = 3SE +/- 809.90, N = 3SE +/- 411.11, N = 3SE +/- 209.97, N = 32500242501622506952510752754722772052772602772982775892776472983412996103008333017403084343098961. (CC) gcc options: -pedantic -O3

FFmpeg

This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 7.0Encoder: libx265 - Scenario: UploadXeon Max 9468 2PEPYC 8534PEPYC 8534PNEPYC 9754 2PXeon Max 9480 2PXeon Max 9468EPYC 9754Xeon Max 9480Xeon Platinum 8490H 2PXeon Platinum 8490HEPYC 9654 2PXeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9684X 2PEPYC 9684XEPYC 9654714212835SE +/- 0.12, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.07, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.07, N = 3SE +/- 0.10, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 322.3423.2523.2723.2823.3023.3323.3623.7724.5624.9726.4126.8026.9327.4327.6127.661. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 7.0Encoder: libx265 - Scenario: Video On DemandXeon Max 9468 2PXeon Max 9480 2PEPYC 8534PEPYC 8534PNEPYC 9754 2PEPYC 9754Xeon Max 9468Xeon Max 9480Xeon Platinum 8490H 2PXeon Platinum 8490HEPYC 9654 2PXeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9684X 2PEPYC 9684XEPYC 96541224364860SE +/- 0.43, N = 3SE +/- 0.13, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.10, N = 3SE +/- 0.07, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.14, N = 3SE +/- 0.03, N = 3SE +/- 0.07, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 345.2946.7946.8846.9347.1447.3347.3448.1448.7749.8452.8153.2953.9955.0155.5655.591. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

Google SynthMark

SynthMark is a cross platform tool for benchmarking CPU performance under a variety of real-time audio workloads. It uses a polyphonic synthesizer model to provide standardized tests for latency, jitter and computational throughput. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgVoices, More Is BetterGoogle SynthMark 20201109Test: VoiceMark_100EPYC 9754 2PEPYC 8534PEPYC 9754EPYC 8534PNXeon Max 9468 2PXeon Max 9468Xeon Platinum 8490HXeon Max 9480Xeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9654 2PEPYC 9684X 2PEPYC 9684XEPYC 9654Xeon Platinum 8592+ 2PXeon Platinum 8592+2004006008001000SE +/- 0.88, N = 3SE +/- 0.14, N = 3SE +/- 0.45, N = 3SE +/- 0.60, N = 3SE +/- 0.30, N = 3SE +/- 0.26, N = 3SE +/- 0.22, N = 3SE +/- 0.24, N = 3SE +/- 0.48, N = 3SE +/- 1.51, N = 3SE +/- 0.56, N = 3SE +/- 0.72, N = 3SE +/- 0.69, N = 3SE +/- 0.55, N = 3SE +/- 0.38, N = 3SE +/- 0.13, N = 3655.05655.24655.80656.25710.92711.10711.38712.01712.09712.70782.32783.20783.42783.78799.49799.721. (CXX) g++ options: -lm -lpthread -std=c++11 -Ofast

FFmpeg

This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 7.0Encoder: libx265 - Scenario: PlatformXeon Max 9468 2PXeon Max 9480 2PEPYC 8534PNEPYC 8534PEPYC 9754 2PEPYC 9754Xeon Max 9468Xeon Max 9480Xeon Platinum 8490H 2PXeon Platinum 8490HEPYC 9654 2PXeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9684X 2PEPYC 9684XEPYC 96541224364860SE +/- 0.09, N = 3SE +/- 0.27, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.08, N = 3SE +/- 0.12, N = 3SE +/- 0.16, N = 3SE +/- 0.01, N = 3SE +/- 0.15, N = 3SE +/- 0.14, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 345.9546.5946.8646.8747.1347.3147.3248.3548.7649.7952.7953.2253.9255.1155.3855.611. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

WebP Image Encode

This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.2.4Encode Settings: Quality 100, Highest CompressionEPYC 8534PEPYC 8534PNEPYC 9754 2PEPYC 9754Xeon Platinum 8490H 2PXeon Platinum 8490HXeon Max 9468 2PXeon Max 9480 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8592+ 2PXeon Platinum 8592+EPYC 9684XEPYC 9654EPYC 9654 2PEPYC 9684X 2P0.81451.6292.44353.2584.0725SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 6SE +/- 0.00, N = 63.033.033.033.043.073.083.093.093.103.103.463.473.613.623.623.621. (CC) gcc options: -fvisibility=hidden -O2 -lm

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PUSCH Processor Benchmark, Throughput ThreadEPYC 8534PEPYC 8534PNEPYC 9754EPYC 9754 2PEPYC 9654EPYC 9654 2PEPYC 9684XEPYC 9684X 2PXeon Platinum 8592+Xeon Platinum 8592+ 2PXeon Max 9468Xeon Max 9468 2PXeon Max 9480Xeon Max 9480 2PXeon Platinum 8490HXeon Platinum 8490H 2P306090120150SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3112.1112.1112.1112.1112.2112.2112.2112.2129.5129.5129.7129.7129.7129.7129.7129.7MIN: 79.3MIN: 79.3MIN: 79.3MIN: 79.3MIN: 79.3MIN: 79.3MIN: 79.3MIN: 79.3MIN: 105.7MIN: 105.7MIN: 105.9MIN: 105.9MIN: 105.9MIN: 105.9MIN: 105.9MIN: 105.91. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -mavx512f -mavx512cd -mavx512bw -mavx512dq -fno-trapping-math -fno-math-errno -ldl

CPU Power Consumption Monitor

OpenBenchmarking.orgWattsCPU Power Consumption MonitorPhoronix Test Suite System MonitoringXeon Max 9480 2PXeon Max 9468 2PEPYC 9684X 2PXeon Platinum 8490H 2PEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9754 2PXeon Max 9480Xeon Max 9468EPYC 9684XXeon Platinum 8490HXeon Platinum 8592+EPYC 9654EPYC 9754EPYC 8534PEPYC 8534PN170340510680850Min: 95.52 / Avg: 569.8 / Max: 760.61Min: 102.49 / Avg: 566.6 / Max: 756.48Min: 33.66 / Avg: 525.49 / Max: 837.69Min: 87.48 / Avg: 517.33 / Max: 758.17Min: 35.7 / Avg: 464.49 / Max: 751.79Min: 39.72 / Avg: 455.07 / Max: 946.96Min: 20.22 / Avg: 375.51 / Max: 735.74Min: 38.16 / Avg: 295.53 / Max: 384.53Min: 38.11 / Avg: 291.53 / Max: 379.76Min: 17.06 / Avg: 290.08 / Max: 413.51Min: 30.38 / Avg: 278.1 / Max: 383.24Min: 17.77 / Avg: 260.08 / Max: 405.42Min: 14.69 / Avg: 244.69 / Max: 373.1Min: 10.09 / Avg: 225.05 / Max: 397.44Min: 8.43 / Avg: 128.7 / Max: 184.62Min: 9.26 / Avg: 114.13 / Max: 155.71

Stockfish

OpenBenchmarking.orgWatts, Fewer Is BetterStockfish 16.1CPU Power Consumption MonitorEPYC 9684X 2PEPYC 9654 2PXeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9684XEPYC 9654Xeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9754EPYC 8534PEPYC 8534PN140280420560700Min: 70.77 / Avg: 766.5 / Max: 822.06Min: 57.4 / Avg: 686.5 / Max: 751.79Min: 195.85 / Avg: 680.45 / Max: 758.53Min: 203.83 / Avg: 679.75 / Max: 756.48Min: 166.29 / Avg: 678.39 / Max: 758.17Min: 78.55 / Avg: 672.28 / Max: 755.5Min: 37.25 / Avg: 640.56 / Max: 735.74Min: 35.5 / Avg: 386.38 / Max: 410.66Min: 30.05 / Avg: 347.3 / Max: 373.1Min: 71.31 / Avg: 341.39 / Max: 379.06Min: 74.29 / Avg: 341.12 / Max: 366.15Min: 56.83 / Avg: 339.56 / Max: 379.91Min: 22.35 / Avg: 337.72 / Max: 370.84Min: 21.51 / Avg: 332.06 / Max: 363.69Min: 17.26 / Avg: 174.36 / Max: 184.62Min: 16.19 / Avg: 145.98 / Max: 155.71

OpenBenchmarking.orgNodes Per Second Per Watt, More Is BetterStockfish 16.1Chess BenchmarkXeon Max 9468Xeon Max 9468 2PXeon Max 9480Xeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8490HXeon Platinum 8592+Xeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9684X 2PEPYC 9654 2PEPYC 9684XEPYC 8534PEPYC 9654EPYC 9754EPYC 8534PN130K260K390K520K650K199243.06202981.45215737.78219573.26276776.39285328.34325017.74331127.60470216.10520707.98540060.41555213.84568916.36569277.48571475.98621037.11

OpenBenchmarking.orgNodes Per Second, More Is BetterStockfish 16.1Chess BenchmarkXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HEPYC 8534PXeon Platinum 8592+Xeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PEPYC 9754EPYC 9654EPYC 9684XXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9654 2PEPYC 9684X 2P90M180M270M360M450MSE +/- 1751369.24, N = 12SE +/- 629927.17, N = 3SE +/- 2147688.02, N = 12SE +/- 1304590.97, N = 3SE +/- 2397455.91, N = 12SE +/- 2966712.90, N = 9SE +/- 4344961.90, N = 10SE +/- 2925715.37, N = 11SE +/- 2148606.38, N = 4SE +/- 4652276.19, N = 12SE +/- 3170744.86, N = 15SE +/- 2268830.04, N = 3SE +/- 2482737.87, N = 3SE +/- 7961094.38, N = 12SE +/- 2356843.91, N = 3SE +/- 919073.22, N = 368019088735922309065815696887479991981441097663451379772031494090821877617691897637991977083502145246592226103783012022323707515303991242231. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fno-peel-loops -fno-tracer -pedantic -O3 -funroll-loops -msse -msse3 -mpopcnt -mavx2 -mbmi -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto-partition=one -flto=jobserver

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Model Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: StandardXeon Max 9468 2PXeon Max 9480 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8490H 2PXeon Platinum 8490HXeon Platinum 8592+Xeon Platinum 8592+ 2PEPYC 8534PEPYC 9754 2PEPYC 8534PNEPYC 9754EPYC 9654 2PEPYC 9654EPYC 9684X 2PEPYC 9684X816243240SE +/- 0.40, N = 3SE +/- 0.33, N = 3SE +/- 0.31, N = 3SE +/- 0.11, N = 3SE +/- 0.12, N = 3SE +/- 0.04, N = 3SE +/- 0.09, N = 3SE +/- 0.09, N = 3SE +/- 0.63, N = 15SE +/- 0.23, N = 15SE +/- 0.61, N = 15SE +/- 0.45, N = 15SE +/- 0.61, N = 15SE +/- 0.50, N = 15SE +/- 0.40, N = 15SE +/- 0.61, N = 1536.2336.0033.9933.3329.7529.4129.0228.8527.9327.9026.8326.4525.3223.9222.7921.581. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: StandardXeon Max 9468 2PXeon Max 9480 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8490H 2PXeon Platinum 8490HXeon Platinum 8592+Xeon Platinum 8592+ 2PEPYC 9754 2PEPYC 8534PEPYC 8534PNEPYC 9754EPYC 9654 2PEPYC 9654EPYC 9684X 2PEPYC 9684X1122334455SE +/- 0.31, N = 3SE +/- 0.26, N = 3SE +/- 0.27, N = 3SE +/- 0.09, N = 3SE +/- 0.13, N = 3SE +/- 0.05, N = 3SE +/- 0.10, N = 3SE +/- 0.11, N = 3SE +/- 0.27, N = 15SE +/- 0.87, N = 15SE +/- 0.82, N = 15SE +/- 0.63, N = 15SE +/- 0.92, N = 15SE +/- 0.91, N = 15SE +/- 0.76, N = 15SE +/- 1.36, N = 1527.6127.7829.4330.0033.6134.0034.4634.6635.8736.0837.5337.9639.8042.0744.0646.881. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInference Time Cost (ms), Fewer Is BetterONNX Runtime 1.17Model: fcn-resnet101-11 - Device: CPU - Executor: StandardEPYC 8534PNEPYC 9654 2PEPYC 8534PEPYC 9754 2PEPYC 9684X 2PEPYC 9654EPYC 9684XEPYC 9754Xeon Max 9480Xeon Max 9468Xeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490HXeon Platinum 8592+Xeon Platinum 8592+ 2PXeon Platinum 8490H 2P50100150200250SE +/- 7.19, N = 15SE +/- 0.35, N = 3SE +/- 6.87, N = 12SE +/- 2.91, N = 3SE +/- 2.18, N = 15SE +/- 4.84, N = 15SE +/- 4.76, N = 12SE +/- 2.29, N = 15SE +/- 6.41, N = 15SE +/- 3.36, N = 15SE +/- 2.59, N = 15SE +/- 5.82, N = 12SE +/- 0.09, N = 3SE +/- 0.71, N = 3SE +/- 0.44, N = 3SE +/- 0.69, N = 15226.63220.35216.70203.57182.27182.02178.03166.79155.79139.10121.78121.67116.32109.67100.7492.351. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

OpenBenchmarking.orgInferences Per Second, More Is BetterONNX Runtime 1.17Model: fcn-resnet101-11 - Device: CPU - Executor: StandardEPYC 8534PNEPYC 9654 2PEPYC 8534PEPYC 9754 2PEPYC 9684X 2PEPYC 9654EPYC 9684XEPYC 9754Xeon Max 9480Xeon Max 9468Xeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490HXeon Platinum 8592+Xeon Platinum 8592+ 2PXeon Platinum 8490H 2P3691215SE +/- 0.11914, N = 15SE +/- 0.00718, N = 3SE +/- 0.14467, N = 12SE +/- 0.07023, N = 3SE +/- 0.06453, N = 15SE +/- 0.15150, N = 15SE +/- 0.16422, N = 12SE +/- 0.08489, N = 15SE +/- 0.22869, N = 15SE +/- 0.15651, N = 15SE +/- 0.16487, N = 15SE +/- 0.34632, N = 12SE +/- 0.00664, N = 3SE +/- 0.05824, N = 3SE +/- 0.04271, N = 3SE +/- 0.07619, N = 154.465334.538174.664914.914205.497115.550185.665086.011646.549657.241878.260168.399628.597199.118859.9268410.836231. (CXX) g++ options: -O3 -march=native -ffunction-sections -fdata-sections -mtune=native -flto=auto -fno-fat-lto-objects -ldl -lrt

srsRAN Project

OpenBenchmarking.orgWatts, Fewer Is BettersrsRAN Project 23.10.1-20240325CPU Power Consumption MonitorEPYC 9684X 2PXeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9684XXeon Max 9480Xeon Platinum 8490HXeon Max 9468EPYC 9654Xeon Platinum 8592+EPYC 9754EPYC 8534PEPYC 8534PN130260390520650Min: 65.52 / Avg: 632.5 / Max: 723.62Min: 189.97 / Avg: 617.8 / Max: 721.97Min: 195.95 / Avg: 604.92 / Max: 745.23Min: 148.45 / Avg: 589.02 / Max: 721.35Min: 55.25 / Avg: 559.64 / Max: 633.66Min: 68.27 / Avg: 537.03 / Max: 710Min: 41.23 / Avg: 514.74 / Max: 612.62Min: 32.26 / Avg: 321.41 / Max: 408.75Min: 74.87 / Avg: 289.12 / Max: 373.87Min: 60.24 / Avg: 283.1 / Max: 369.61Min: 74.5 / Avg: 281.29 / Max: 379.76Min: 28.12 / Avg: 276.24 / Max: 341.73Min: 25.9 / Avg: 266.9 / Max: 379.06Min: 21.57 / Avg: 261.04 / Max: 306.84Min: 17.85 / Avg: 136.58 / Max: 166.81Min: 16.71 / Avg: 117.93 / Max: 142.26

OpenBenchmarking.orgMbps Per Watt, More Is BettersrsRAN Project 23.10.1-20240325Test: PDSCH Processor Benchmark, Throughput TotalXeon Max 9480 2PEPYC 9654 2PEPYC 9684X 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9480Xeon Platinum 8592+ 2PEPYC 9754 2PXeon Max 9468Xeon Platinum 8490HEPYC 9684XEPYC 9754Xeon Platinum 8592+EPYC 9654EPYC 8534PEPYC 8534PN408012016020019.2140.9841.0749.9851.1252.1656.8858.0475.6887.2191.4296.5298.83100.26141.87165.23

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PDSCH Processor Benchmark, Throughput TotalXeon Max 9480 2PXeon Max 9480EPYC 8534PEPYC 8534PNXeon Max 9468EPYC 9654 2PXeon Platinum 8490HEPYC 9754EPYC 9684X 2PXeon Platinum 8592+EPYC 9654EPYC 9684XEPYC 9754 2PXeon Platinum 8490H 2PXeon Max 9468 2PXeon Platinum 8592+ 2P7K14K21K28K35KSE +/- 490.54, N = 15SE +/- 281.72, N = 15SE +/- 207.19, N = 5SE +/- 58.21, N = 3SE +/- 938.95, N = 15SE +/- 526.01, N = 15SE +/- 236.27, N = 4SE +/- 237.63, N = 3SE +/- 343.97, N = 3SE +/- 298.34, N = 15SE +/- 357.13, N = 12SE +/- 348.28, N = 4SE +/- 2115.57, N = 15SE +/- 312.17, N = 3SE +/- 175.78, N = 4SE +/- 356.81, N = 1511866.715081.019375.919485.821286.622931.824688.525194.625977.426378.127695.229384.929877.130108.730236.030547.61. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -mavx512f -mavx512cd -mavx512bw -mavx512dq -fno-trapping-math -fno-math-errno -ldl

SVT-AV1

OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 2.0CPU Power Consumption MonitorXeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9684X 2PEPYC 9654 2PXeon Max 9468Xeon Max 9480EPYC 9754 2PXeon Platinum 8490HXeon Platinum 8592+EPYC 9684XEPYC 9654EPYC 9754EPYC 8534PEPYC 8534PN100200300400500Min: 193.69 / Avg: 401.02 / Max: 565.3Min: 196.97 / Avg: 398.41 / Max: 555.31Min: 160.44 / Avg: 317.58 / Max: 460.42Min: 70.69 / Avg: 247.31 / Max: 459.08Min: 60.69 / Avg: 231.16 / Max: 380.35Min: 56.44 / Avg: 206.94 / Max: 347.98Min: 73.18 / Avg: 187.24 / Max: 299.74Min: 69.14 / Avg: 187.17 / Max: 307.31Min: 38.31 / Avg: 171.31 / Max: 283.81Min: 59.06 / Avg: 155.98 / Max: 273.08Min: 26.49 / Avg: 131.08 / Max: 275.44Min: 31.67 / Avg: 118.98 / Max: 227.28Min: 25.79 / Avg: 105.15 / Max: 208.71Min: 19.23 / Avg: 84.14 / Max: 169.24Min: 16.06 / Avg: 60.52 / Max: 126.9Min: 14.1 / Avg: 55.32 / Max: 107.14

OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 13 - Input: Bosphorus 4KXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9684X 2PXeon Max 9468Xeon Max 9480EPYC 9654 2PEPYC 9754 2PXeon Platinum 8490HXeon Platinum 8592+EPYC 9684XEPYC 9654EPYC 9754EPYC 8534PEPYC 8534PN0.62391.24781.87172.49563.11950.3080.3100.4860.6590.6980.7730.8010.8080.9331.0421.3341.3661.6441.9422.6572.773

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 13 - Input: Bosphorus 4KXeon Max 9468 2PXeon Max 9480 2PXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490H 2PEPYC 9754 2PEPYC 8534PEPYC 9684X 2PXeon Platinum 8490HEPYC 9684XXeon Platinum 8592+ 2PEPYC 9754EPYC 9654 2PEPYC 9654Xeon Platinum 8592+4080120160200SE +/- 1.63, N = 15SE +/- 1.67, N = 15SE +/- 1.44, N = 15SE +/- 1.62, N = 15SE +/- 3.84, N = 15SE +/- 2.18, N = 15SE +/- 4.46, N = 15SE +/- 4.25, N = 15SE +/- 1.54, N = 15SE +/- 2.31, N = 15SE +/- 3.81, N = 15SE +/- 2.84, N = 15SE +/- 4.48, N = 15SE +/- 4.23, N = 15SE +/- 4.38, N = 15SE +/- 2.83, N = 15122.62124.26144.70150.00153.42154.50159.75160.79161.34162.47162.48162.98163.41167.22172.88174.831. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

LuxCoreRender

OpenBenchmarking.orgWatts, Fewer Is BetterLuxCoreRender 2.6CPU Power Consumption MonitorEPYC 9684X 2PXeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9754 2PEPYC 9684XXeon Max 9468Xeon Platinum 8490HXeon Max 9480Xeon Platinum 8592+EPYC 9754EPYC 9654EPYC 8534PEPYC 8534PN150300450600750Min: 65.89 / Avg: 739.45 / Max: 837.69Min: 203.55 / Avg: 668.4 / Max: 754.98Min: 195.7 / Avg: 665.76 / Max: 753.2Min: 173.98 / Avg: 664.56 / Max: 753.01Min: 76.68 / Avg: 654.92 / Max: 755.35Min: 59.66 / Avg: 573.93 / Max: 661.76Min: 42.95 / Avg: 561.1 / Max: 675.69Min: 34.75 / Avg: 384.12 / Max: 412.45Min: 74.56 / Avg: 331.26 / Max: 374.04Min: 59.67 / Avg: 330.53 / Max: 373.83Min: 72.86 / Avg: 330.29 / Max: 377.65Min: 24.28 / Avg: 326.43 / Max: 375.69Min: 20.73 / Avg: 311.89 / Max: 340.37Min: 29.9 / Avg: 299.39 / Max: 326.71Min: 18.62 / Avg: 156.14 / Max: 171.55Min: 18.19 / Avg: 128.95 / Max: 140.58

OpenBenchmarking.orgM samples/sec Per Watt, More Is BetterLuxCoreRender 2.6Scene: DLSC - Acceleration: CPUXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2PEPYC 9684XEPYC 9654EPYC 9754EPYC 8534PEPYC 8534PN0.01490.02980.04470.05960.07450.0150.0160.0170.0190.0230.0240.0260.0290.0310.0310.0390.0420.0480.0540.0600.066

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: DLSC - Acceleration: CPUXeon Max 9468Xeon Max 9480EPYC 8534PNXeon Platinum 8490HXeon Platinum 8592+EPYC 8534PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9684XEPYC 9754EPYC 9654 2PEPYC 9754 2PEPYC 9684X 2P510152025SE +/- 0.08, N = 3SE +/- 0.10, N = 3SE +/- 0.03, N = 3SE +/- 0.06, N = 15SE +/- 0.02, N = 3SE +/- 0.06, N = 15SE +/- 0.12, N = 3SE +/- 0.20, N = 15SE +/- 0.23, N = 15SE +/- 0.24, N = 15SE +/- 0.13, N = 3SE +/- 0.18, N = 15SE +/- 0.28, N = 15SE +/- 0.12, N = 3SE +/- 1.02, N = 15SE +/- 0.91, N = 157.468.008.488.729.389.3910.1710.4411.1212.7414.2916.2116.7917.8922.0022.58MIN: 7.17 / MAX: 8.21MIN: 7.65 / MAX: 9.12MIN: 8.26 / MAX: 9.41MIN: 8.12 / MAX: 10.6MIN: 8.94 / MAX: 11.49MIN: 8.9 / MAX: 10.42MIN: 9.34 / MAX: 14.75MIN: 8.86 / MAX: 16.08MIN: 9.18 / MAX: 17.51MIN: 10.69 / MAX: 19.19MIN: 13.78 / MAX: 17.04MIN: 15.2 / MAX: 19.13MIN: 15.23 / MAX: 21.36MIN: 16.29 / MAX: 28.22MIN: 16.32 / MAX: 35.07MIN: 17.46 / MAX: 32.78

NAS Parallel Benchmarks

OpenBenchmarking.orgWatts, Fewer Is BetterNAS Parallel Benchmarks 3.4CPU Power Consumption MonitorXeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PEPYC 9684X 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9754 2PEPYC 9684XXeon Platinum 8490HXeon Platinum 8592+Xeon Max 9480Xeon Max 9468EPYC 9654EPYC 9754EPYC 8534PEPYC 8534PN140280420560700Min: 196.31 / Avg: 561.84 / Max: 754.16Min: 200.97 / Avg: 537.97 / Max: 722.6Min: 155.56 / Avg: 520.43 / Max: 752.57Min: 65.8 / Avg: 495.73 / Max: 819.01Min: 68.69 / Avg: 471.7 / Max: 753.27Min: 56.44 / Avg: 416.97 / Max: 683.16Min: 40.5 / Avg: 379.03 / Max: 677.9Min: 31.93 / Avg: 307.77 / Max: 409.54Min: 57.58 / Avg: 291.31 / Max: 376.98Min: 26.35 / Avg: 276.68 / Max: 375.06Min: 76.21 / Avg: 270.13 / Max: 345.76Min: 73.35 / Avg: 254.67 / Max: 351.04Min: 26.47 / Avg: 242.89 / Max: 334.43Min: 21.19 / Avg: 234.01 / Max: 336.9Min: 18.8 / Avg: 129.08 / Max: 158.73Min: 17.1 / Avg: 113.27 / Max: 135.62

OpenBenchmarking.orgTotal Mop/s Per Watt, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: EP.DXeon Max 9468 2PXeon Max 9468Xeon Max 9480 2PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8490H 2PXeon Platinum 8592+Xeon Platinum 8592+ 2PEPYC 9684XEPYC 9654EPYC 9684X 2PEPYC 9654 2PEPYC 8534PEPYC 8534PNEPYC 9754EPYC 9754 2P132639526520.3921.0722.7723.8625.9628.7229.2433.9337.1243.8644.7246.5649.7150.7253.7458.09

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: EP.DXeon Max 9468EPYC 8534PNEPYC 8534PXeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654Xeon Max 9468 2PEPYC 9684XEPYC 9754Xeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9754 2PEPYC 9684X 2P5K10K15K20K25KSE +/- 155.10, N = 12SE +/- 50.01, N = 8SE +/- 14.68, N = 3SE +/- 6.53, N = 3SE +/- 6.77, N = 3SE +/- 22.33, N = 3SE +/- 122.47, N = 15SE +/- 4.83, N = 4SE +/- 309.03, N = 15SE +/- 367.88, N = 12SE +/- 10.93, N = 4SE +/- 100.23, N = 4SE +/- 30.19, N = 4SE +/- 530.31, N = 15SE +/- 1274.85, N = 12SE +/- 602.37, N = 155366.695745.046416.346445.097562.808089.1410652.0310970.4511424.0512575.3012793.0114946.7616004.0719414.9422017.7522166.601. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6

Pennant

OpenBenchmarking.orgWatts, Fewer Is BetterPennant 1.0.1CPU Power Consumption MonitorXeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PEPYC 9684X 2PXeon Platinum 8592+ 2PEPYC 9654 2PEPYC 9754 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HEPYC 9684XXeon Platinum 8592+EPYC 9654EPYC 9754EPYC 8534PEPYC 8534PN140280420560700Min: 191.18 / Avg: 450.17 / Max: 753.29Min: 187.85 / Avg: 448.22 / Max: 752.01Min: 144.69 / Avg: 414.55 / Max: 753.5Min: 63.33 / Avg: 370.49 / Max: 818.97Min: 64.47 / Avg: 342.65 / Max: 753.04Min: 54.46 / Avg: 322.18 / Max: 670.5Min: 41.42 / Avg: 291.74 / Max: 665.7Min: 71.8 / Avg: 241.62 / Max: 375.61Min: 74.45 / Avg: 234.93 / Max: 375.6Min: 57.31 / Avg: 222.1 / Max: 377.34Min: 33.15 / Avg: 196.35 / Max: 409.53Min: 25.17 / Avg: 191.11 / Max: 377.76Min: 27.07 / Avg: 170.76 / Max: 330.73Min: 20.01 / Avg: 156.67 / Max: 328.66Min: 18.14 / Avg: 107.18 / Max: 164.25Min: 16.11 / Avg: 95.67 / Max: 145.23

OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: leblancbigEPYC 8534PNEPYC 8534PXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9654EPYC 9754EPYC 9684XXeon Max 9468 2PXeon Platinum 8490H 2PXeon Max 9480 2PEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9684X 2P246810SE +/- 0.045713, N = 13SE +/- 0.057580, N = 6SE +/- 0.025580, N = 6SE +/- 0.023904, N = 7SE +/- 0.022393, N = 7SE +/- 0.022698, N = 7SE +/- 0.115723, N = 15SE +/- 0.115702, N = 15SE +/- 0.115724, N = 15SE +/- 0.010204, N = 7SE +/- 0.019069, N = 15SE +/- 0.008354, N = 7SE +/- 0.063684, N = 15SE +/- 0.013568, N = 6SE +/- 0.022246, N = 15SE +/- 0.055800, N = 156.8527126.5417685.9309575.1141614.9558654.1827073.7548353.6543033.3604532.8810992.5298072.5273632.3129772.1594592.0843842.0768351. (CXX) g++ options: -fopenmp -lmpi_cxx -lmpi

NAMD

OpenBenchmarking.orgWatts, Fewer Is BetterNAMD 3.0b6CPU Power Consumption MonitorXeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9684X 2PXeon Max 9480EPYC 9654 2PXeon Max 9468EPYC 9684XXeon Platinum 8490HXeon Platinum 8592+EPYC 9754 2PEPYC 9654EPYC 9754EPYC 8534PEPYC 8534PN140280420560700Min: 194.25 / Avg: 565.07 / Max: 686.05Min: 192.28 / Avg: 560.13 / Max: 714.79Min: 163.14 / Avg: 552.42 / Max: 653.46Min: 73.13 / Avg: 510.74 / Max: 686.96Min: 65.49 / Avg: 343.59 / Max: 811.63Min: 72.24 / Avg: 295.54 / Max: 371.41Min: 57.51 / Avg: 289.28 / Max: 672.46Min: 71.79 / Avg: 281.33 / Max: 364.12Min: 32.28 / Avg: 274.76 / Max: 406.72Min: 55.23 / Avg: 273.91 / Max: 344.33Min: 33.09 / Avg: 260.62 / Max: 349.3Min: 42.36 / Avg: 249.94 / Max: 651.86Min: 28.55 / Avg: 235.77 / Max: 330Min: 21.48 / Avg: 206.2 / Max: 309.47Min: 18.33 / Avg: 130.18 / Max: 164.25Min: 15.81 / Avg: 114.62 / Max: 140.3

OpenBenchmarking.orgns/day Per Watt, More Is BetterNAMD 3.0b6Input: STMV with 1,066,628 AtomsXeon Platinum 8490H 2PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490HXeon Platinum 8592+ 2PXeon Max 9468Xeon Max 9480Xeon Platinum 8592+EPYC 8534PEPYC 9654EPYC 9684XEPYC 8534PNEPYC 9754EPYC 9654 2PEPYC 9684X 2PEPYC 9754 2P0.00520.01040.01560.02080.0260.0020.0030.0030.0040.0040.0050.0050.0060.0120.0120.0120.0130.0140.0180.0190.023

OpenBenchmarking.orgns/day, More Is BetterNAMD 3.0b6Input: STMV with 1,066,628 AtomsXeon Platinum 8490HXeon Platinum 8490H 2PXeon Platinum 8592+Xeon Max 9468EPYC 8534PNXeon Max 9480EPYC 8534PXeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8592+ 2PEPYC 9654EPYC 9754EPYC 9684XEPYC 9654 2PEPYC 9754 2PEPYC 9684X 2P246810SE +/- 0.01396, N = 15SE +/- 0.04403, N = 15SE +/- 0.02648, N = 15SE +/- 0.05087, N = 15SE +/- 0.00540, N = 3SE +/- 0.05862, N = 15SE +/- 0.00440, N = 3SE +/- 0.05463, N = 15SE +/- 0.05644, N = 12SE +/- 0.08185, N = 13SE +/- 0.00537, N = 3SE +/- 0.00847, N = 3SE +/- 0.00206, N = 3SE +/- 0.02927, N = 3SE +/- 0.00332, N = 3SE +/- 0.01739, N = 31.052741.109001.437871.452231.461631.502141.519011.767781.886792.254222.878772.888693.410855.278395.836066.38343

OpenBenchmarking.orgWatts, Fewer Is BetterNAMD 3.0b6CPU Power Consumption MonitorXeon Max 9480 2PXeon Platinum 8490H 2PXeon Max 9468 2PXeon Platinum 8592+ 2PEPYC 9684X 2PXeon Max 9480EPYC 9654 2PXeon Platinum 8490HXeon Max 9468Xeon Platinum 8592+EPYC 9754 2PEPYC 9654EPYC 9684XEPYC 9754EPYC 8534PEPYC 8534PN140280420560700Min: 190.7 / Avg: 550.3 / Max: 686.19Min: 165.98 / Avg: 533.68 / Max: 633.47Min: 194.24 / Avg: 522.05 / Max: 693.75Min: 75.47 / Avg: 484.77 / Max: 671.95Min: 62.44 / Avg: 326.5 / Max: 779.77Min: 73.99 / Avg: 269.88 / Max: 376.76Min: 58.84 / Avg: 262.91 / Max: 648.62Min: 55.01 / Avg: 258.53 / Max: 352.06Min: 73.54 / Avg: 248.99 / Max: 374.54Min: 33.34 / Avg: 236.13 / Max: 373.62Min: 40.88 / Avg: 225.05 / Max: 615.88Min: 27.01 / Avg: 187.68 / Max: 312.41Min: 29.28 / Avg: 180.85 / Max: 379.02Min: 20.68 / Avg: 165.67 / Max: 289.24Min: 17 / Avg: 109.51 / Max: 166.07Min: 16.45 / Avg: 96.77 / Max: 141.98

OpenBenchmarking.orgns/day Per Watt, More Is BetterNAMD 3.0b6Input: ATPase with 327,506 AtomsXeon Platinum 8490H 2PXeon Max 9480 2PXeon Platinum 8490HXeon Platinum 8592+ 2PXeon Max 9468 2PXeon Platinum 8592+Xeon Max 9480Xeon Max 9468EPYC 9654EPYC 9754EPYC 8534PEPYC 9684XEPYC 8534PNEPYC 9684X 2PEPYC 9654 2PEPYC 9754 2P0.01890.03780.05670.07560.09450.0040.0050.0110.0120.0140.0200.0220.0240.0440.0480.0520.0520.0560.0640.0680.084

OpenBenchmarking.orgns/day, More Is BetterNAMD 3.0b6Input: ATPase with 327,506 AtomsXeon Platinum 8490H 2PXeon Max 9480 2PXeon Platinum 8490HXeon Platinum 8592+EPYC 8534PNEPYC 8534PXeon Platinum 8592+ 2PXeon Max 9480Xeon Max 9468Xeon Max 9468 2PEPYC 9754EPYC 9654EPYC 9684XEPYC 9654 2PEPYC 9754 2PEPYC 9684X 2P510152025SE +/- 0.17435, N = 15SE +/- 0.44989, N = 15SE +/- 0.21567, N = 12SE +/- 0.54361, N = 15SE +/- 0.00431, N = 4SE +/- 0.04338, N = 5SE +/- 0.34017, N = 12SE +/- 0.14537, N = 15SE +/- 0.16058, N = 15SE +/- 0.26203, N = 15SE +/- 0.04917, N = 5SE +/- 0.01368, N = 5SE +/- 0.02148, N = 6SE +/- 0.04751, N = 3SE +/- 0.02125, N = 3SE +/- 0.08771, N = 32.219482.563192.970214.648145.449715.706725.852675.925015.971307.300657.997568.182759.3353417.8110518.9745420.75259

miniFE

OpenBenchmarking.orgWatts, Fewer Is BetterminiFE 2.2CPU Power Consumption MonitorXeon Max 9480 2PXeon Max 9468 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9684X 2PEPYC 9654 2PXeon Max 9480EPYC 9754 2PXeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+EPYC 9684XEPYC 9654EPYC 9754EPYC 8534PEPYC 8534PN130260390520650Min: 199.32 / Avg: 500.61 / Max: 723.88Min: 189.32 / Avg: 498.77 / Max: 730.29Min: 168.8 / Avg: 424.3 / Max: 732.03Min: 73.08 / Avg: 366.83 / Max: 700.4Min: 63.04 / Avg: 330.55 / Max: 718.24Min: 53.09 / Avg: 309.09 / Max: 579.88Min: 73.91 / Avg: 260.35 / Max: 366.7Min: 40.05 / Avg: 258.88 / Max: 518.97Min: 75.77 / Avg: 253.53 / Max: 365.64Min: 61.92 / Avg: 224.98 / Max: 377.92Min: 31.62 / Avg: 198.71 / Max: 331.13Min: 33.7 / Avg: 177.88 / Max: 364.34Min: 28.79 / Avg: 152.24 / Max: 300.13Min: 21.36 / Avg: 127.63 / Max: 261.53Min: 19.07 / Avg: 96.36 / Max: 147.72Min: 15.16 / Avg: 86.83 / Max: 132.07

OpenBenchmarking.orgCG Mflops Per Watt, More Is BetterminiFE 2.2Problem Size: SmallXeon Max 9468 2PXeon Max 9480 2PXeon Max 9480Xeon Max 9468Xeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9654 2PXeon Platinum 8490HXeon Platinum 8592+EPYC 9754 2PEPYC 9684X 2PEPYC 8534PEPYC 8534PNEPYC 9684XEPYC 9654EPYC 97549018027036045039.8543.1171.5279.24110.48114.81140.84156.59160.18162.67179.90254.52282.72298.90334.41392.87

OpenBenchmarking.orgCG Mflops, More Is BetterminiFE 2.2Problem Size: SmallXeon Max 9480Xeon Max 9468 2PXeon Max 9468Xeon Max 9480 2PEPYC 8534PEPYC 8534PNXeon Platinum 8592+Xeon Platinum 8490HEPYC 9754 2PXeon Platinum 8592+ 2PEPYC 9654 2PXeon Platinum 8490H 2PEPYC 9754EPYC 9654EPYC 9684XEPYC 9684X 2P13K26K39K52K65KSE +/- 572.00, N = 12SE +/- 1097.75, N = 15SE +/- 601.03, N = 12SE +/- 283.44, N = 3SE +/- 49.20, N = 4SE +/- 42.17, N = 4SE +/- 135.77, N = 4SE +/- 11.14, N = 4SE +/- 2087.13, N = 15SE +/- 1633.92, N = 15SE +/- 1510.37, N = 15SE +/- 467.78, N = 4SE +/- 53.07, N = 5SE +/- 42.89, N = 5SE +/- 92.04, N = 5SE +/- 2182.93, N = 1518621.219876.520091.121583.224526.024548.831830.235230.142113.542114.643531.046877.550139.950909.653167.859466.61. (CXX) g++ options: -O3 -fopenmp -lmpi_cxx -lmpi

RocksDB

OpenBenchmarking.orgWatts, Fewer Is BetterRocksDB 9.0CPU Power Consumption MonitorEPYC 9684X 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Max 9468 2PEPYC 9654 2PXeon Platinum 8592+ 2PEPYC 9754 2PEPYC 9684XXeon Max 9468Xeon Max 9480Xeon Platinum 8490HXeon Platinum 8592+EPYC 9754EPYC 9654EPYC 8534PEPYC 8534PN140280420560700Min: 67.1 / Avg: 671.28 / Max: 820.11Min: 185.7 / Avg: 663.3 / Max: 752.58Min: 157.04 / Avg: 656.04 / Max: 755.91Min: 200.41 / Avg: 651.38 / Max: 755.47Min: 56.63 / Avg: 576.34 / Max: 690.47Min: 67.94 / Avg: 558.28 / Max: 752.93Min: 43.23 / Avg: 505.19 / Max: 682.09Min: 34.3 / Avg: 383.63 / Max: 411.02Min: 72.31 / Avg: 332.58 / Max: 376.04Min: 74.89 / Avg: 331.82 / Max: 375.82Min: 52.65 / Avg: 330.54 / Max: 376.66Min: 33.11 / Avg: 323.9 / Max: 378.14Min: 22 / Avg: 301.4 / Max: 339.84Min: 24.6 / Avg: 300.55 / Max: 337.2Min: 17.34 / Avg: 155.39 / Max: 168.8Min: 17.36 / Avg: 134 / Max: 149.2

OpenBenchmarking.orgOp/s Per Watt, More Is BetterRocksDB 9.0Test: Read While WritingXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Max 9480Xeon Platinum 8592+ 2PXeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+EPYC 9684X 2PEPYC 9654 2PEPYC 9684XEPYC 9754 2PEPYC 9654EPYC 9754EPYC 8534PEPYC 8534PN14K28K42K56K70K14775.4916920.4618781.2722200.6222256.9522369.2623723.7627560.2829572.8934889.6536906.8345330.1848605.6952342.4660067.2166672.70

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Read While WritingXeon Max 9480Xeon Max 9468Xeon Platinum 8490HXeon Platinum 8592+EPYC 8534PNEPYC 8534PXeon Max 9468 2PXeon Max 9480 2PXeon Platinum 8490H 2PXeon Platinum 8592+ 2PEPYC 9684XEPYC 9654EPYC 9754EPYC 9684X 2PEPYC 9654 2PEPYC 9754 2P5M10M15M20M25MSE +/- 26369.56, N = 3SE +/- 83772.27, N = 13SE +/- 61518.27, N = 10SE +/- 28201.44, N = 3SE +/- 81998.56, N = 3SE +/- 28098.62, N = 3SE +/- 218479.19, N = 15SE +/- 254193.37, N = 15SE +/- 343737.14, N = 15SE +/- 215352.04, N = 15SE +/- 93669.97, N = 15SE +/- 36524.26, N = 3SE +/- 109424.14, N = 3SE +/- 236709.61, N = 3SE +/- 506458.47, N = 15SE +/- 273109.47, N = 373665007439663784173889268938934176933364596244171122332912321300124255191415841914608544157761861985166420108289229001521. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

259 Results Shown

OpenVINO
BRL-CAD
ASKAP:
  tConvolve MPI - Degridding
  tConvolve MPI - Gridding
NAS Parallel Benchmarks
OpenFOAM
OpenSSL
Pennant
Xmrig
OpenSSL
7-Zip Compression
John The Ripper:
  Blowfish
  bcrypt
OpenRadioss
Primesieve
OpenVINO
Llamafile
OpenSSL
Xcompact3d Incompact3d
OpenSSL
OpenVINO:
  Weld Porosity Detection FP16-INT8 - CPU
  Person Detection FP16 - CPU
Speedb
RocksDB
OpenVINO:
  Person Re-Identification Retail FP16 - CPU
  Road Segmentation ADAS FP16-INT8 - CPU
John The Ripper
OpenSSL
RELION
OpenVINO
QuantLib
m-queens
Coremark
Blender
PostgreSQL
Primesieve
PostgreSQL
OpenSSL
ONNX Runtime
OSPRay Studio
Blender:
  Pabellon Barcelona - CPU-Only
  Barbershop - CPU-Only
OSPRay Studio
PyTorch
OpenSSL:
  AES-256-GCM
  AES-128-GCM
OpenVINO
NAS Parallel Benchmarks
OpenVINO
Helsing
OpenVINO
PyTorch
OSPRay Studio
PyTorch
OpenVINO
OSPRay Studio
NAS Parallel Benchmarks:
  CG.C
  IS.D
ASTC Encoder
ASKAP
Blender:
  BMW27 - CPU-Only
  Fishy Cat - CPU-Only
miniBUDE:
  OpenMP - BM2:
    Billion Interactions/s
    GFInst/s
OpenVINO
ASTC Encoder
ACES DGEMM
OSPRay Studio:
  1 - 4K - 1 - Path Tracer - CPU
  3 - 4K - 1 - Path Tracer - CPU
LULESH
Speedb
PyTorch:
  CPU - 256 - ResNet-152
  CPU - 512 - ResNet-152
OpenVINO
PyTorch
NAS Parallel Benchmarks
Algebraic Multi-Grid Benchmark
srsRAN Project
ONNX Runtime
OpenVINO
Aircrack-ng
Blender
OSPRay
ONNX Runtime
WRF
miniBUDE:
  OpenMP - BM1:
    Billion Interactions/s
    GFInst/s
ONNX Runtime
Embree
ONNX Runtime:
  T5 Encoder - CPU - Standard
  ResNet50 v1-12-int8 - CPU - Parallel
GROMACS
Kvazaar
ONNX Runtime
Llamafile
OSPRay:
  particle_volume/ao/real_time
  particle_volume/scivis/real_time
Kvazaar
OpenVKL
Xcompact3d Incompact3d
OpenVINO
Liquid-DSP
Embree
ASTC Encoder
uvg266
Liquid-DSP
Embree
High Performance Conjugate Gradient
ONNX Runtime:
  GPT-2 - CPU - Standard
  CaffeNet 12-int8 - CPU - Parallel
Intel Open Image Denoise
ONNX Runtime
Intel Open Image Denoise
OpenVINO
uvg266
Intel Open Image Denoise
GraphicsMagick
OpenVINO
oneDNN
ONNX Runtime
Liquid-DSP
ONNX Runtime
LuxCoreRender
GPAW
OpenVINO
ONNX Runtime
Memcached
OpenFOAM
OpenVINO:
  Age Gender Recognition Retail 0013 FP16-INT8 - CPU
  Person Re-Identification Retail FP16 - CPU
TensorFlow
GraphicsMagick
LAMMPS Molecular Dynamics Simulator
CloverLeaf
Timed Node.js Compilation
ONNX Runtime
7-Zip Compression
ONNX Runtime
Parallel BZIP2 Compression
LAMMPS Molecular Dynamics Simulator
OpenVINO
Timed LLVM Compilation
PostgreSQL:
  100 - 1000 - Read Write - Average Latency
  100 - 1000 - Read Write
Timed Linux Kernel Compilation
ONNX Runtime
OpenRadioss
CloverLeaf
TensorFlow
ASTC Encoder
Y-Cruncher
TensorFlow
OpenVINO
Y-Cruncher
Kvazaar
Liquid-DSP
OpenVINO
TensorFlow
LuxCoreRender
GNU Octave Benchmark
GraphicsMagick
ONNX Runtime
Liquid-DSP
TensorFlow:
  CPU - 256 - GoogLeNet
  CPU - 512 - ResNet-50
Apache IoTDB
TensorFlow
Apache IoTDB
Timed Linux Kernel Compilation
uvg266
Liquid-DSP
ClickHouse:
  100M Rows Hits Dataset, Third Run
  100M Rows Hits Dataset, Second Run
NWChem
OpenRadioss
ClickHouse
TensorFlow
Liquid-DSP
Timed Godot Game Engine Compilation
QMCPACK
Timed FFmpeg Compilation
Apache IoTDB
OpenFOAM
GraphicsMagick
TensorFlow
Liquid-DSP
SVT-AV1
Timed Gem5 Compilation
Timed ImageMagick Compilation
Apache IoTDB
Liquid-DSP
Appleseed
Zstd Compression
OpenVINO
SVT-AV1
Zstd Compression
x265
SVT-AV1
VVenC
OpenFOAM
PyBench
Liquid-DSP:
  64 - 256 - 57
  64 - 256 - 32
QuantLib
ONNX Runtime
libavif avifenc
srsRAN Project
Zstd Compression
Liquid-DSP
Zstd Compression
Timed Mesa Compilation
libavif avifenc
Numpy Benchmark
libavif avifenc:
  0
  10, Lossless
  2
VVenC
WebP Image Encode
SecureMark
FFmpeg:
  libx265 - Upload
  libx265 - Video On Demand
Google SynthMark
FFmpeg
WebP Image Encode
srsRAN Project
CPU Power Consumption Monitor:
  Phoronix Test Suite System Monitoring
  CPU Power Consumption Monitor
  Chess Benchmark
Stockfish
ONNX Runtime:
  Faster R-CNN R-50-FPN-int8 - CPU - Standard:
    Inference Time Cost (ms)
    Inferences Per Second
  fcn-resnet101-11 - CPU - Standard:
    Inference Time Cost (ms)
    Inferences Per Second
srsRAN Project:
  CPU Power Consumption Monitor
  PDSCH Processor Benchmark, Throughput Total
srsRAN Project
SVT-AV1:
  CPU Power Consumption Monitor
  Preset 13 - Bosphorus 4K
SVT-AV1
LuxCoreRender:
  CPU Power Consumption Monitor
  DLSC - CPU
LuxCoreRender
NAS Parallel Benchmarks:
  CPU Power Consumption Monitor
  EP.D
NAS Parallel Benchmarks
Pennant
Pennant
NAMD:
  CPU Power Consumption Monitor
  STMV with 1,066,628 Atoms
NAMD
NAMD:
  CPU Power Consumption Monitor
  ATPase with 327,506 Atoms
NAMD
miniFE:
  CPU Power Consumption Monitor
  Small
miniFE
RocksDB:
  CPU Power Consumption Monitor
  Read While Writing
RocksDB