OpenVINO AMX Benchmarks AMD EPYC Genoa

Benchmarks by Michael Larabel for a future article.

HTML result view exported from: https://openbenchmarking.org/result/2301166-NE-OPENVINOA64&grw&rdt.

OpenVINO AMX Benchmarks AMD EPYC GenoaProcessorMotherboardChipsetMemoryDiskGraphicsMonitorNetworkOSKernelDesktopDisplay ServerVulkanCompilerFile-SystemScreen ResolutionXeon Platinum 8490H 2PXeon Platinum 8490H 2P - No AMXEPYC 9654 2P2 x Intel Xeon Platinum 8490H @ 3.50GHz (120 Cores / 240 Threads)Quanta Cloud S6Q-MB-MPS (3A10.uh BIOS)Intel Device 1bce16 x 64 GB 4800MT/s Samsung M321R8GA0BB0-CQKEG2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007 + 960GB INTEL SSDSC2KG96ASPEEDVGA HDMI4 x Intel E810-C for QSFP + 2 x Intel X710 for 10GBASE-TUbuntu 22.046.1.4-060104-generic (x86_64)GNOME Shell 42.2X Server 1.21.1.31.2.204GCC 11.3.0 + Clang 14.0.0-1ubuntu1ext41920x10802 x AMD EPYC 9654 96-Core @ 2.40GHz (192 Cores / 384 Threads)AMD Titanite_4G (RTI1002E BIOS)AMD Device 14a424 x 64 GB 4800MT/s Samsung M321R8GA0BB0-CQKEG2 x 1920GB SAMSUNG MZWLJ1T9HBJR-00007Broadcom NetXtreme BCM5720 PCIeOpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseCompiler Details- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-xKiWfi/gcc-11-11.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-xKiWfi/gcc-11-11.3.0/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-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Processor Details- Xeon Platinum 8490H 2P: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b0000c0- Xeon Platinum 8490H 2P - No AMX: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b0000c0- EPYC 9654 2P: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa10110dPython Details- Python 3.10.6Security Details- Xeon Platinum 8490H 2P: 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 2P - No AMX: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9654 2P: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

OpenVINO AMX Benchmarks AMD EPYC Genoaopenvino: Face Detection FP16-INT8 - CPUopenvino: Face Detection FP16-INT8 - CPUopenvino: Weld Porosity Detection FP16-INT8 - CPUopenvino: Weld Porosity Detection FP16-INT8 - CPUopenvino: Person Vehicle Bike Detection FP16 - CPUopenvino: Person Vehicle Bike Detection FP16 - CPUopenvino: Machine Translation EN To DE FP16 - CPUopenvino: Machine Translation EN To DE FP16 - CPUXeon Platinum 8490H 2PXeon Platinum 8490H 2P - No AMXEPYC 9654 2P282.25424.1623066.635.184778.3425.09724.1082.79146.08409.818897.1013.462730.0721.95321.28186.53188.64507.2518678.1610.265320.8318.03989.1697.00OpenBenchmarking.org

OpenVINO

Model: Face Detection FP16-INT8 - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Face Detection FP16-INT8 - Device: CPUXeon Platinum 8490H 2PXeon Platinum 8490H 2P - No AMXEPYC 9654 2P60120180240300SE +/- 2.56, N = 3SE +/- 0.29, N = 3SE +/- 0.17, N = 3282.25146.08188.641. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie -ldl

OpenVINO

Model: Face Detection FP16-INT8 - Device: CPU

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Face Detection FP16-INT8 - Device: CPUXeon Platinum 8490H 2PXeon Platinum 8490H 2P - No AMXEPYC 9654 2P110220330440550SE +/- 3.75, N = 3SE +/- 0.81, N = 3SE +/- 0.31, N = 3424.16409.81507.25MIN: 153.43 / MAX: 1952.95MIN: 203.37 / MAX: 909.92MIN: 231.44 / MAX: 623.691. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie -ldl

OpenVINO

Model: Weld Porosity Detection FP16-INT8 - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Weld Porosity Detection FP16-INT8 - Device: CPUXeon Platinum 8490H 2PXeon Platinum 8490H 2P - No AMXEPYC 9654 2P5K10K15K20K25KSE +/- 92.31, N = 3SE +/- 24.48, N = 3SE +/- 10.84, N = 323066.638897.1018678.161. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie -ldl

OpenVINO

Model: Weld Porosity Detection FP16-INT8 - Device: CPU

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Weld Porosity Detection FP16-INT8 - Device: CPUXeon Platinum 8490H 2PXeon Platinum 8490H 2P - No AMXEPYC 9654 2P3691215SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 35.1813.4610.26MIN: 1.43 / MAX: 102.62MIN: 3.82 / MAX: 88.11MIN: 4.21 / MAX: 95.331. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie -ldl

OpenVINO

Model: Person Vehicle Bike Detection FP16 - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Person Vehicle Bike Detection FP16 - Device: CPUXeon Platinum 8490H 2PXeon Platinum 8490H 2P - No AMXEPYC 9654 2P11002200330044005500SE +/- 14.65, N = 3SE +/- 11.17, N = 3SE +/- 44.61, N = 84778.342730.075320.831. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie -ldl

OpenVINO

Model: Person Vehicle Bike Detection FP16 - Device: CPU

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Person Vehicle Bike Detection FP16 - Device: CPUXeon Platinum 8490H 2PXeon Platinum 8490H 2P - No AMXEPYC 9654 2P612182430SE +/- 0.08, N = 3SE +/- 0.09, N = 3SE +/- 0.15, N = 825.0921.9518.03MIN: 11.66 / MAX: 132.09MIN: 11.62 / MAX: 120.04MIN: 5.34 / MAX: 112.511. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie -ldl

OpenVINO

Model: Machine Translation EN To DE FP16 - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.3Model: Machine Translation EN To DE FP16 - Device: CPUXeon Platinum 8490H 2PXeon Platinum 8490H 2P - No AMXEPYC 9654 2P2004006008001000SE +/- 5.21, N = 3SE +/- 0.77, N = 3SE +/- 8.13, N = 8724.10321.28989.161. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie -ldl

OpenVINO

Model: Machine Translation EN To DE FP16 - Device: CPU

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.3Model: Machine Translation EN To DE FP16 - Device: CPUXeon Platinum 8490H 2PXeon Platinum 8490H 2P - No AMXEPYC 9654 2P4080120160200SE +/- 0.60, N = 3SE +/- 0.45, N = 3SE +/- 0.78, N = 882.79186.5397.00MIN: 31.64 / MAX: 1588.97MIN: 92.82 / MAX: 808.7MIN: 41.26 / MAX: 1899.381. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie -ldl

OpenVINO

CPU Peak Freq (Highest CPU Core Frequency) Monitor

MinAvgMaxXeon Platinum 8490H 2P190025613507Xeon Platinum 8490H 2P - No AMX123123743508EPYC 9654 2P240029903702OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2022.3CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000

OpenVINO

CPU Power Consumption Monitor

MinAvgMaxXeon Platinum 8490H 2P208616761Xeon Platinum 8490H 2P - No AMX210641748EPYC 9654 2P54602717OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption Monitor2004006008001000

OpenVINO

CPU Peak Freq (Highest CPU Core Frequency) Monitor

MinAvgMaxXeon Platinum 8490H 2P190028603501Xeon Platinum 8490H 2P - No AMX190029283512EPYC 9654 2P240028093702OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2022.3CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000

OpenVINO

CPU Power Consumption Monitor

MinAvgMaxXeon Platinum 8490H 2P217663757Xeon Platinum 8490H 2P - No AMX210665756EPYC 9654 2P59652717OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption Monitor2004006008001000

OpenVINO

CPU Peak Freq (Highest CPU Core Frequency) Monitor

MinAvgMaxXeon Platinum 8490H 2P190029653501Xeon Platinum 8490H 2P - No AMX190027393511EPYC 9654 2P240034933763OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2022.3CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000

OpenVINO

CPU Power Consumption Monitor

MinAvgMaxXeon Platinum 8490H 2P213640755Xeon Platinum 8490H 2P - No AMX132653756EPYC 9654 2P55567643OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption Monitor2004006008001000

OpenVINO

CPU Peak Freq (Highest CPU Core Frequency) Monitor

MinAvgMaxXeon Platinum 8490H 2P190027923516Xeon Platinum 8490H 2P - No AMX190026563508EPYC 9654 2P240029653700OpenBenchmarking.orgMegahertz, More Is BetterOpenVINO 2022.3CPU Peak Freq (Highest CPU Core Frequency) Monitor10002000300040005000

OpenVINO

CPU Power Consumption Monitor

MinAvgMaxXeon Platinum 8490H 2P212644742Xeon Platinum 8490H 2P - No AMX211652756EPYC 9654 2P55650740OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2022.3CPU Power Consumption Monitor2004006008001000

CPU Peak Freq (Highest CPU Core Frequency) Monitor

Phoronix Test Suite System Monitoring

OpenBenchmarking.orgMegahertzCPU Peak Freq (Highest CPU Core Frequency) MonitorPhoronix Test Suite System MonitoringXeon Platinum 8490H 2PXeon Platinum 8490H 2P - No AMXEPYC 9654 2P7001400210028003500Min: 1900 / Avg: 2744.33 / Max: 3516Min: 1000 / Avg: 2603.67 / Max: 3512Min: 2400 / Avg: 3129.46 / Max: 3763

CPU Power Consumption Monitor

Phoronix Test Suite System Monitoring

OpenBenchmarking.orgWattsCPU Power Consumption MonitorPhoronix Test Suite System MonitoringXeon Platinum 8490H 2PXeon Platinum 8490H 2P - No AMXEPYC 9654 2P130260390520650Min: 208.07 / Avg: 624.35 / Max: 760.79Min: 122.95 / Avg: 646.09 / Max: 763.9Min: 53.42 / Avg: 596.25 / Max: 739.81

Phoronix Test Suite v10.8.4