Intel Xeon Platinum 8490H Linux Benchmarks

Tests for a future article on Phoronix of Sapphire Rapids launch day review using the Intel Xeon Platinum 8490H processors. More benchmarks to come by Michael Larabel.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2301109-PTS-SPRREVIE33
Jump To Table - Results

View

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

Limit displaying results to tests within:

BLAS (Basic Linear Algebra Sub-Routine) Tests 2 Tests
C++ Boost Tests 3 Tests
Chess Test Suite 2 Tests
Timed Code Compilation 4 Tests
C/C++ Compiler Tests 7 Tests
CPU Massive 20 Tests
Creator Workloads 14 Tests
Fortran Tests 4 Tests
Game Development 4 Tests
HPC - High Performance Computing 19 Tests
Linear Algebra 2 Tests
Machine Learning 4 Tests
Molecular Dynamics 6 Tests
MPI Benchmarks 5 Tests
Multi-Core 25 Tests
NVIDIA GPU Compute 4 Tests
Intel oneAPI 6 Tests
OpenMPI Tests 13 Tests
Programmer / Developer System Benchmarks 9 Tests
Python 2 Tests
Raytracing 5 Tests
Renderers 6 Tests
Scientific Computing 8 Tests
Server 4 Tests
Server CPU Tests 15 Tests
Single-Threaded 3 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

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
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
EPYC 7713
September 10 2022
  11 Hours, 38 Minutes
EPYC 7713 2P
September 06 2022
  8 Hours, 48 Minutes
EPYC 7763
September 16 2022
  11 Hours, 27 Minutes
EPYC 7763 2P
September 18 2022
  8 Hours, 42 Minutes
EPYC 7773X
September 19 2022
  10 Hours, 45 Minutes
EPYC 7773X 2P
September 22 2022
  8 Hours, 48 Minutes
EPYC 9374F
November 11 2022
  10 Hours, 13 Minutes
EPYC 9374F 2P
November 13 2022
  7 Hours, 50 Minutes
EPYC 9554
November 02 2022
  9 Hours
EPYC 9554 2P
October 31 2022
  7 Hours, 15 Minutes
EPYC 9654
October 15 2022
  8 Hours, 29 Minutes
EPYC 9654 2P
October 22 2022
  6 Hours, 42 Minutes
Xeon Platinum 8362
October 04 2022
  13 Hours, 52 Minutes
Xeon Platinum 8362 2P
October 05 2022
  9 Hours, 22 Minutes
Xeon Platinum 8380
September 30 2022
  12 Hours, 57 Minutes
Xeon Platinum 8380 2P
November 04 2022
  9 Hours, 53 Minutes
Xeon Platinum 8490H
January 08 2023
  10 Hours, 20 Minutes
Xeon Platinum 8490H 2P
January 06 2023
  8 Hours, 52 Minutes
Invert Hiding All Results Option
  9 Hours, 43 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):


Intel Xeon Platinum 8490H Linux BenchmarksProcessorMotherboardChipsetMemoryDiskGraphicsMonitorNetworkOSKernelDesktopDisplay ServerVulkanCompilerFile-SystemScreen ResolutionXeon Platinum 8490HXeon Platinum 8490H 2PIntel Xeon Platinum 8490H @ 3.50GHz (60 Cores / 120 Threads)Quanta Cloud S6Q-MB-MPS (3A10.uh BIOS)Intel Device 1bce512GB3841GB Micron_9300_MTFDHAL3T8TDPASPEEDVGA HDMI4 x Intel E810-C for QSFPUbuntu 22.106.0.0-060000rc3daily20220904-generic (x86_64)GNOME ShellX Server 1.21.1.31.3.211GCC 12.2.0ext41920x10802 x Intel Xeon Platinum 8490H @ 3.50GHz (120 Cores / 240 Threads)1008GB4 x Intel E810-C for QSFP + 2 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-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Processor Details- Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b0000c0Java Details- OpenJDK Runtime Environment (build 11.0.16+8-post-Ubuntu-0ubuntu1)Python Details- Python 3.10.6Security Details- itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected

Xeon Platinum 8490H vs. Xeon Platinum 8490H 2P ComparisonPhoronix Test SuiteBaseline+47.5%+47.5%+95%+95%+142.5%+142.5%189.9%180.3%180.3%170.3%149.7%106%98.2%94.9%94.8%94.2%89%86.3%83.8%83.5%60.6%60.3%57.1%57%56.8%56.8%49.6%48.9%46.1%45.7%37.8%13.9%6.7%W.P.D.F - CPUF.D.F - CPUF.D.F - CPUW.P.D.F.I - CPUF.D.F.I - CPUP.V.B.D.F - CPUC.B.S.A - u8s8f32 - CPUN.D.C.o.b.u.o.I - A.M.SC.C.R.5.I - A.M.SN.T.C.B.b.u.c - A.M.SN.Q.A.B.b.u.S.1.P - A.M.SC.D.Y.C - A.M.SC.B.S.A - u8s8f32 - CPUA.G.R.R.0.F - CPUM.T.E.T.D.F - CPUM.T.E.T.D.F - CPUP.D.F - CPUP.D.F - CPUP.D.F - CPUP.D.F - CPUC.B.S.A - f32 - CPUD.B.s - f32 - CPUC.B.S.A - f32 - CPUW.P.D.F - CPUA.G.R.R.0.F - CPUEigenC.B.S.A - bf16bf16bf16 - CPUOpenVINOOpenVINOOpenVINOOpenVINOOpenVINOOpenVINOoneDNNNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseoneDNNOpenVINOOpenVINOOpenVINOOpenVINOOpenVINOOpenVINOOpenVINOoneDNNoneDNNoneDNNOpenVINOOpenVINOLeelaChessZerooneDNNXeon Platinum 8490HXeon Platinum 8490H 2P

Intel Xeon Platinum 8490H Linux Benchmarksopenvino: Face Detection FP16 - CPUopenvino: Face Detection FP16 - CPUopenvino: Face Detection FP16-INT8 - CPUopenvino: Age Gender Recognition Retail 0013 FP16 - CPUopenvino: Age Gender Recognition Retail 0013 FP16 - CPUopenvino: Person Detection FP16 - CPUopenvino: Person Detection FP16 - CPUopenvino: Person Detection FP32 - CPUopenvino: Person Detection FP32 - CPUopenvino: Weld Porosity Detection FP16-INT8 - CPUopenvino: Weld Porosity Detection FP16 - CPUopenvino: Weld Porosity Detection FP16 - CPUopenvino: Person Vehicle Bike Detection FP16 - CPUopenvino: Machine Translation EN To DE FP16 - CPUopenvino: Machine Translation EN To DE FP16 - CPUlczero: Eigenonednn: Convolution Batch Shapes Auto - f32 - CPUonednn: Convolution Batch Shapes Auto - u8s8f32 - CPUonednn: Deconvolution Batch shapes_3d - f32 - CPUonednn: Convolution Batch Shapes Auto - f32 - CPUonednn: Convolution Batch Shapes Auto - u8s8f32 - CPUonednn: Convolution Batch Shapes Auto - bf16bf16bf16 - CPUdeepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Streamdeepsparse: CV Detection,YOLOv5s COCO - Asynchronous Multi-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-StreamXeon Platinum 8490HXeon Platinum 8490H 2P70.58424.67150.8782929.060.6225.551167.8325.531169.1516703.647839.517.523656.94448.5966.66105130.6363940.4925161.108120.6457790.5194510.281812279.45953.5734769.779347.4454192.9925318.932547.2877197.81151.49376.67152169.330.4540.14744.6540.08745.8645149.7522725.255.167533.49719.2241.51119760.4356540.2679390.7440210.4317880.2621140.264160281.81553.54341499.500492.1557364.7952594.022992.1671OpenBenchmarking.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 2022.2.devModel: Face Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P4080120160200SE +/- 0.05, N = 3SE +/- 0.12, N = 370.58197.811. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Face Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P4080120160200Min: 70.52 / Avg: 70.58 / Max: 70.68Min: 197.63 / Avg: 197.81 / Max: 198.051. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Face Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P90180270360450SE +/- 0.29, N = 3SE +/- 0.09, N = 3424.67151.49MIN: 324.89 / MAX: 491.29MIN: 123.14 / MAX: 273.71. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Face Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P80160240320400Min: 424.09 / Avg: 424.67 / Max: 425.03Min: 151.33 / Avg: 151.49 / Max: 151.621. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Face Detection FP16-INT8 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P80160240320400SE +/- 0.45, N = 3SE +/- 0.66, N = 3150.87376.671. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Face Detection FP16-INT8 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P70140210280350Min: 149.97 / Avg: 150.87 / Max: 151.32Min: 375.55 / Avg: 376.67 / Max: 377.851. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Age Gender Recognition Retail 0013 FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P30K60K90K120K150KSE +/- 111.96, N = 3SE +/- 667.44, N = 382929.06152169.331. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Age Gender Recognition Retail 0013 FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P30K60K90K120K150KMin: 82810.79 / Avg: 82929.06 / Max: 83152.86Min: 151131.4 / Avg: 152169.33 / Max: 153415.241. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Age Gender Recognition Retail 0013 FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P0.13950.2790.41850.5580.6975SE +/- 0.00, N = 3SE +/- 0.00, N = 30.620.45MIN: 0.31 / MAX: 23.57MIN: 0.33 / MAX: 78.341. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Age Gender Recognition Retail 0013 FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P246810Min: 0.62 / Avg: 0.62 / Max: 0.62Min: 0.45 / Avg: 0.45 / Max: 0.451. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Person Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P918273645SE +/- 0.05, N = 3SE +/- 0.10, N = 325.5540.141. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Person Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P816243240Min: 25.45 / Avg: 25.55 / Max: 25.62Min: 39.97 / Avg: 40.14 / Max: 40.331. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Person Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P30060090012001500SE +/- 2.53, N = 3SE +/- 2.09, N = 31167.83744.65MIN: 600.36 / MAX: 1656.84MIN: 515.46 / MAX: 1463.521. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Person Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P2004006008001000Min: 1164.07 / Avg: 1167.83 / Max: 1172.64Min: 740.98 / Avg: 744.65 / Max: 748.21. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Person Detection FP32 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P918273645SE +/- 0.01, N = 3SE +/- 0.45, N = 325.5340.081. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Person Detection FP32 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P816243240Min: 25.51 / Avg: 25.53 / Max: 25.55Min: 39.34 / Avg: 40.08 / Max: 40.91. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Person Detection FP32 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P30060090012001500SE +/- 0.26, N = 3SE +/- 8.26, N = 31169.15745.86MIN: 870.58 / MAX: 1651.76MIN: 475.57 / MAX: 1408.181. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Person Detection FP32 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P2004006008001000Min: 1168.66 / Avg: 1169.15 / Max: 1169.52Min: 730.98 / Avg: 745.86 / Max: 759.51. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Weld Porosity Detection FP16-INT8 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P10K20K30K40K50KSE +/- 9.01, N = 3SE +/- 54.08, N = 316703.6445149.751. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Weld Porosity Detection FP16-INT8 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P8K16K24K32K40KMin: 16691.57 / Avg: 16703.64 / Max: 16721.26Min: 45046.81 / Avg: 45149.75 / Max: 45229.981. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Weld Porosity Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P5K10K15K20K25KSE +/- 60.72, N = 10SE +/- 252.27, N = 47839.5122725.251. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Weld Porosity Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P4K8K12K16K20KMin: 7710.6 / Avg: 7839.51 / Max: 8370.38Min: 22085.37 / Avg: 22725.25 / Max: 23206.261. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Weld Porosity Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P246810SE +/- 0.05, N = 10SE +/- 0.05, N = 47.525.16MIN: 2.66 / MAX: 31.82MIN: 4.27 / MAX: 91.311. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Weld Porosity Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P3691215Min: 7.1 / Avg: 7.52 / Max: 7.63Min: 5.07 / Avg: 5.16 / Max: 5.281. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Person Vehicle Bike Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P16003200480064008000SE +/- 5.03, N = 3SE +/- 17.07, N = 33656.947533.491. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Person Vehicle Bike Detection FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P13002600390052006500Min: 3649.83 / Avg: 3656.94 / Max: 3666.67Min: 7502.46 / Avg: 7533.49 / Max: 7561.341. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Machine Translation EN To DE FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P160320480640800SE +/- 0.03, N = 3SE +/- 2.72, N = 3448.59719.221. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Machine Translation EN To DE FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P130260390520650Min: 448.54 / Avg: 448.59 / Max: 448.63Min: 714.22 / Avg: 719.22 / Max: 723.591. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Machine Translation EN To DE FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P1530456075SE +/- 0.01, N = 3SE +/- 0.15, N = 366.6641.51MIN: 23.93 / MAX: 106.71MIN: 27.82 / MAX: 338.881. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Machine Translation EN To DE FP16 - Device: CPUXeon Platinum 8490HXeon Platinum 8490H 2P1326395265Min: 66.64 / Avg: 66.66 / Max: 66.68Min: 41.27 / Avg: 41.51 / Max: 41.791. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -flto -shared

LeelaChessZero

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

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.28Backend: EigenXeon Platinum 8490HXeon Platinum 8490H 2P3K6K9K12K15KSE +/- 64.57, N = 3SE +/- 76.29, N = 310513119761. (CXX) g++ options: -flto -pthread
OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.28Backend: EigenXeon Platinum 8490HXeon Platinum 8490H 2P2K4K6K8K10KMin: 10415 / Avg: 10513.33 / Max: 10635Min: 11855 / Avg: 11976 / Max: 121171. (CXX) g++ options: -flto -pthread

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 Intel oneAPI. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPUXeon Platinum 8490HXeon Platinum 8490H 2P0.14320.28640.42960.57280.716SE +/- 0.000967, N = 7SE +/- 0.002329, N = 70.6363940.435654MIN: 0.54MIN: 0.381. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPUXeon Platinum 8490HXeon Platinum 8490H 2P246810Min: 0.63 / Avg: 0.64 / Max: 0.64Min: 0.43 / Avg: 0.44 / Max: 0.441. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPUXeon Platinum 8490HXeon Platinum 8490H 2P0.11080.22160.33240.44320.554SE +/- 0.000571, N = 7SE +/- 0.001562, N = 70.4925160.267939MIN: 0.42MIN: 0.21. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPUXeon Platinum 8490HXeon Platinum 8490H 2P246810Min: 0.49 / Avg: 0.49 / Max: 0.5Min: 0.26 / Avg: 0.27 / Max: 0.271. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPUXeon Platinum 8490HXeon Platinum 8490H 2P0.24930.49860.74790.99721.2465SE +/- 0.000893, N = 9SE +/- 0.001670, N = 91.1081200.744021MIN: 1.08MIN: 0.671. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.6Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPUXeon Platinum 8490HXeon Platinum 8490H 2P246810Min: 1.11 / Avg: 1.11 / Max: 1.11Min: 0.74 / Avg: 0.74 / Max: 0.751. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -std=c++11 -pie -ldl -lpthread

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPUXeon Platinum 8490HXeon Platinum 8490H 2P0.14530.29060.43590.58120.7265SE +/- 0.000493, N = 7SE +/- 0.002510, N = 70.6457790.431788MIN: 0.55MIN: 0.381. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPUXeon Platinum 8490HXeon Platinum 8490H 2P246810Min: 0.64 / Avg: 0.65 / Max: 0.65Min: 0.42 / Avg: 0.43 / Max: 0.441. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPUXeon Platinum 8490HXeon Platinum 8490H 2P0.11690.23380.35070.46760.5845SE +/- 0.001949, N = 7SE +/- 0.002009, N = 100.5194510.262114MIN: 0.46MIN: 0.21. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPUXeon Platinum 8490HXeon Platinum 8490H 2P246810Min: 0.51 / Avg: 0.52 / Max: 0.53Min: 0.26 / Avg: 0.26 / Max: 0.281. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPUXeon Platinum 8490HXeon Platinum 8490H 2P0.06340.12680.19020.25360.317SE +/- 0.002592, N = 15SE +/- 0.001768, N = 70.2818120.264160MIN: 0.25MIN: 0.221. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread
OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPUXeon Platinum 8490HXeon Platinum 8490H 2P12345Min: 0.27 / Avg: 0.28 / Max: 0.3Min: 0.26 / Avg: 0.26 / Max: 0.271. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl -lpthread

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-StreamXeon Platinum 8490HXeon Platinum 8490H 2P60120180240300SE +/- 0.50, N = 3SE +/- 0.80, N = 3279.46281.82
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-StreamXeon Platinum 8490HXeon Platinum 8490H 2P50100150200250Min: 278.57 / Avg: 279.46 / Max: 280.29Min: 280.4 / Avg: 281.82 / Max: 283.17

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-StreamXeon Platinum 8490HXeon Platinum 8490H 2P0.8041.6082.4123.2164.02SE +/- 0.0063, N = 3SE +/- 0.0099, N = 33.57343.5434
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-StreamXeon Platinum 8490HXeon Platinum 8490H 2P246810Min: 3.56 / Avg: 3.57 / Max: 3.58Min: 3.53 / Avg: 3.54 / Max: 3.56

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamXeon Platinum 8490HXeon Platinum 8490H 2P30060090012001500SE +/- 0.49, N = 3SE +/- 2.06, N = 3769.781499.50
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamXeon Platinum 8490HXeon Platinum 8490H 2P30060090012001500Min: 769.05 / Avg: 769.78 / Max: 770.71Min: 1495.37 / Avg: 1499.5 / Max: 1501.67

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamXeon Platinum 8490HXeon Platinum 8490H 2P20406080100SE +/- 0.06, N = 3SE +/- 0.09, N = 347.4592.16
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamXeon Platinum 8490HXeon Platinum 8490H 2P20406080100Min: 47.34 / Avg: 47.45 / Max: 47.55Min: 92.02 / Avg: 92.16 / Max: 92.32

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-StreamXeon Platinum 8490HXeon Platinum 8490H 2P80160240320400SE +/- 0.12, N = 3SE +/- 2.04, N = 3192.99364.80
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-StreamXeon Platinum 8490HXeon Platinum 8490H 2P70140210280350Min: 192.86 / Avg: 192.99 / Max: 193.23Min: 360.96 / Avg: 364.8 / Max: 367.91

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Detection,YOLOv5s COCO - Scenario: Asynchronous Multi-StreamXeon Platinum 8490HXeon Platinum 8490H 2P130260390520650SE +/- 0.14, N = 3SE +/- 0.43, N = 3318.93594.02
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: CV Detection,YOLOv5s COCO - Scenario: Asynchronous Multi-StreamXeon Platinum 8490HXeon Platinum 8490H 2P100200300400500Min: 318.66 / Avg: 318.93 / Max: 319.1Min: 593.41 / Avg: 594.02 / Max: 594.86

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamXeon Platinum 8490HXeon Platinum 8490H 2P20406080100SE +/- 0.15, N = 3SE +/- 0.24, N = 347.2992.17
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.1Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamXeon Platinum 8490HXeon Platinum 8490H 2P20406080100Min: 47.02 / Avg: 47.29 / Max: 47.55Min: 91.91 / Avg: 92.17 / Max: 92.64

Geometric Mean Of All Test Results

OpenBenchmarking.orgGeometric Mean, More Is BetterGeometric Mean Of All Test ResultsResult Composite - Intel Xeon Platinum 8490H Linux BenchmarksXeon Platinum 8490HXeon Platinum 8490H 2P4080120160200110.54190.28

30 Results Shown

OpenVINO:
  Face Detection FP16 - CPU:
    FPS
    ms
  Face Detection FP16-INT8 - CPU:
    FPS
  Age Gender Recognition Retail 0013 FP16 - CPU:
    FPS
    ms
  Person Detection FP16 - CPU:
    FPS
    ms
  Person Detection FP32 - CPU:
    FPS
    ms
  Weld Porosity Detection FP16-INT8 - CPU:
    FPS
  Weld Porosity Detection FP16 - CPU:
    FPS
    ms
  Person Vehicle Bike Detection FP16 - CPU:
    FPS
  Machine Translation EN To DE FP16 - CPU:
    FPS
    ms
LeelaChessZero
oneDNN:
  Convolution Batch Shapes Auto - f32 - CPU
  Convolution Batch Shapes Auto - u8s8f32 - CPU
  Deconvolution Batch shapes_3d - f32 - CPU
oneDNN:
  Convolution Batch Shapes Auto - f32 - CPU
  Convolution Batch Shapes Auto - u8s8f32 - CPU
  Convolution Batch Shapes Auto - bf16bf16bf16 - CPU
Neural Magic DeepSparse:
  CV Classification, ResNet-50 ImageNet - Synchronous Single-Stream:
    items/sec
    ms/batch
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream:
    items/sec
  NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream:
    items/sec
  NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Stream:
    items/sec
  CV Detection,YOLOv5s COCO - Asynchronous Multi-Stream:
    items/sec
  NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream:
    items/sec
Geometric Mean Of All Test Results