AMD EPYC PlaidML

AMD EPYC PlaidML.

HTML result view exported from: https://openbenchmarking.org/result/2009172-FI-PLAIDING711.

AMD EPYC PlaidMLProcessorMotherboardChipsetMemoryDiskGraphicsMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverCompilerFile-SystemScreen ResolutionOpenGL2 x EPYC 72522 x EPYC 72622 x EPYC 72722 x EPYC 74522 x EPYC 77422 x EPYC 7F322 x EPYC 7F722 x Xeon Gold 6230R2 x Xeon Silver 4210REPYC 7232PEPYC 7302PEPYC 7402PEPYC 7502PXeon Platinum 82802 x AMD EPYC 7252 8-Core @ 3.10GHz (16 Cores / 32 Threads)Supermicro H11DSi-NT v2.00 (2.1 BIOS)AMD Starship/Matisse504GB280GB INTEL SSDPE21D280GAASPEEDVE2282 x Intel 10G X550TUbuntu 20.045.8.0-050800rc6daily20200721-generic (x86_64) 20200720GNOME Shell 3.36.1X Server 1.20.8modesetting 1.20.8GCC 9.3.0ext41920x10802 x AMD EPYC 7262 8-Core @ 3.20GHz (16 Cores / 32 Threads)2 x AMD EPYC 7272 12-Core @ 2.90GHz (24 Cores / 48 Threads)2 x AMD EPYC 7452 32-Core @ 2.35GHz (64 Cores / 128 Threads)llvmpipe 504GBVGA HDMI3.3 Mesa 20.0.4 (LLVM 9.0.1 128 bits)2 x AMD EPYC 7742 64-Core @ 2.25GHz (128 Cores / 256 Threads)2 x AMD EPYC 7F32 8-Core @ 3.70GHz (16 Cores / 32 Threads)llvmpipe2 x AMD EPYC 7F72 24-Core @ 3.20GHz (48 Cores / 96 Threads)2 x Intel Xeon Gold 6230R @ 4.00GHz (52 Cores / 104 Threads)GIGABYTE MD61-SC2-00 v01000100 (T15 BIOS)Intel Sky Lake-E DMI3 Registers378GBASPEEDVE2282 x Intel X722 for 1GbE + 2 x QLogic FastLinQ QL41000 10/25/40/50GbE2 x Intel Xeon Silver 4210R @ 3.20GHz (20 Cores / 40 Threads)AMD EPYC 7232P 8-Core @ 3.10GHz (8 Cores / 16 Threads)Supermicro H11DSi-NT v2.00 (2.1 BIOS)AMD Starship/Matisse252GB2 x Intel 10G X550TAMD EPYC 7302P 16-Core @ 3.00GHz (16 Cores / 32 Threads)AMD EPYC 7402P 24-Core @ 2.80GHz (24 Cores / 48 Threads)AMD EPYC 7502P 32-Core @ 2.50GHz (32 Cores / 64 Threads)Intel Xeon Platinum 8280 @ 4.00GHz (28 Cores / 56 Threads)GIGABYTE MD61-SC2-00 v01000100 (T15 BIOS)Intel Sky Lake-E DMI3 Registers188GB2 x Intel X722 for 1GbE + 2 x QLogic FastLinQ QL41000 10/25/40/50GbE1024x768OpenBenchmarking.orgProcessor Details- 2 x EPYC 7252: Scaling Governor: acpi-cpufreq performance - CPU Microcode: 0x8301034- 2 x EPYC 7262: Scaling Governor: acpi-cpufreq performance - CPU Microcode: 0x8301034- 2 x EPYC 7272: Scaling Governor: acpi-cpufreq performance - CPU Microcode: 0x8301034- 2 x EPYC 7452: Scaling Governor: acpi-cpufreq performance - CPU Microcode: 0x8301034- 2 x EPYC 7742: Scaling Governor: acpi-cpufreq performance - CPU Microcode: 0x8301034- 2 x EPYC 7F32: Scaling Governor: acpi-cpufreq performance - CPU Microcode: 0x8301034- 2 x EPYC 7F72: Scaling Governor: acpi-cpufreq performance - CPU Microcode: 0x8301034- 2 x Xeon Gold 6230R: Scaling Governor: intel_pstate performance - CPU Microcode: 0x500002c- 2 x Xeon Silver 4210R: Scaling Governor: intel_pstate performance - CPU Microcode: 0x500002c- EPYC 7232P: Scaling Governor: acpi-cpufreq performance - CPU Microcode: 0x8301034- EPYC 7302P: Scaling Governor: acpi-cpufreq performance - CPU Microcode: 0x8301034- EPYC 7402P: Scaling Governor: acpi-cpufreq performance - CPU Microcode: 0x8301034- EPYC 7502P: Scaling Governor: acpi-cpufreq performance - CPU Microcode: 0x8301034- Xeon Platinum 8280: Scaling Governor: intel_pstate performance - CPU Microcode: 0x500002cPython Details- Python 2.7.18rc1 + Python 3.8.2Security Details- 2 x EPYC 7252: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected- 2 x EPYC 7262: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected- 2 x EPYC 7272: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected- 2 x EPYC 7452: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected- 2 x EPYC 7742: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected- 2 x EPYC 7F32: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected- 2 x EPYC 7F72: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected- 2 x Xeon Gold 6230R: itlb_multihit: KVM: Mitigation of Split huge pages + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Mitigation of TSX disabled- 2 x Xeon Silver 4210R: itlb_multihit: KVM: Mitigation of Split huge pages + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Mitigation of TSX disabled- EPYC 7232P: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected- EPYC 7302P: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected- EPYC 7402P: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected- EPYC 7502P: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected- Xeon Platinum 8280: itlb_multihit: KVM: Mitigation of Split huge pages + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Mitigation of TSX disabled

AMD EPYC PlaidMLplaidml: No - Inference - VGG16 - CPUplaidml: No - Inference - VGG19 - CPUplaidml: No - Inference - IMDB LSTM - CPUplaidml: No - Inference - Mobilenet - CPUplaidml: No - Inference - ResNet 50 - CPUplaidml: No - Inference - DenseNet 201 - CPUplaidml: No - Inference - Inception V3 - CPU2 x EPYC 72522 x EPYC 72622 x EPYC 72722 x EPYC 74522 x EPYC 77422 x EPYC 7F322 x EPYC 7F722 x Xeon Gold 6230R2 x Xeon Silver 4210REPYC 7232PEPYC 7302PEPYC 7402PEPYC 7502PXeon Platinum 828020.1616.33428.6011.324.452.708.1820.1716.55388.4811.314.532.737.9723.4619.48435.5611.574.512.718.5527.8724.50462.4511.464.452.678.2229.3725.20334.9811.384.502.597.7822.4918.73419.1512.214.822.998.6727.9424.23449.5211.934.762.888.7525.0822.73644.5112.014.7915.3112.70506.368.453.651.634.6811.839.431197.2612.185.172.627.7822.5118.20615.1913.615.6739.9727.3722.76634.2814.245.773.0010.8632.1926.611019.4914.545.973.0711.0331.9126.61637.8415.186.282.577.35OpenBenchmarking.org

PlaidML

FP16: No - Mode: Inference - Network: VGG16 - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: VGG16 - Device: CPU2 x EPYC 72522 x EPYC 72622 x EPYC 72722 x EPYC 74522 x EPYC 77422 x EPYC 7F322 x EPYC 7F722 x Xeon Gold 6230R2 x Xeon Silver 4210REPYC 7232PEPYC 7302PEPYC 7402PEPYC 7502PXeon Platinum 8280714212835SE +/- 0.16, N = 3SE +/- 0.05, N = 3SE +/- 0.21, N = 10SE +/- 0.43, N = 3SE +/- 0.69, N = 12SE +/- 0.14, N = 3SE +/- 0.32, N = 15SE +/- 0.25, N = 3SE +/- 0.11, N = 3SE +/- 0.08, N = 3SE +/- 0.26, N = 3SE +/- 0.25, N = 3SE +/- 0.22, N = 3SE +/- 0.23, N = 320.1620.1723.4627.8729.3722.4927.9425.0815.3111.8322.5127.3732.1931.91

PlaidML

FP16: No - Mode: Inference - Network: VGG19 - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: VGG19 - Device: CPU2 x EPYC 72522 x EPYC 72622 x EPYC 72722 x EPYC 74522 x EPYC 77422 x EPYC 7F322 x EPYC 7F722 x Xeon Gold 6230R2 x Xeon Silver 4210REPYC 7232PEPYC 7302PEPYC 7402PEPYC 7502PXeon Platinum 8280612182430SE +/- 0.24, N = 4SE +/- 0.08, N = 3SE +/- 0.23, N = 3SE +/- 0.22, N = 15SE +/- 0.29, N = 15SE +/- 0.11, N = 3SE +/- 0.03, N = 3SE +/- 0.32, N = 3SE +/- 0.04, N = 3SE +/- 0.10, N = 3SE +/- 0.21, N = 3SE +/- 0.19, N = 3SE +/- 0.19, N = 3SE +/- 0.15, N = 316.3316.5519.4824.5025.2018.7324.2322.7312.709.4318.2022.7626.6126.61

PlaidML

FP16: No - Mode: Inference - Network: IMDB LSTM - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: IMDB LSTM - Device: CPU2 x EPYC 72522 x EPYC 72622 x EPYC 72722 x EPYC 74522 x EPYC 77422 x EPYC 7F322 x EPYC 7F722 x Xeon Gold 6230R2 x Xeon Silver 4210REPYC 7232PEPYC 7302PEPYC 7402PEPYC 7502PXeon Platinum 828030060090012001500SE +/- 4.43, N = 3SE +/- 3.80, N = 3SE +/- 4.83, N = 3SE +/- 4.33, N = 3SE +/- 2.84, N = 15SE +/- 5.82, N = 15SE +/- 2.83, N = 3SE +/- 18.26, N = 15SE +/- 1.28, N = 3SE +/- 4.55, N = 3SE +/- 2.80, N = 3SE +/- 1.99, N = 3SE +/- 16.96, N = 3SE +/- 2.58, N = 3428.60388.48435.56462.45334.98419.15449.52644.51506.361197.26615.19634.281019.49637.84

PlaidML

FP16: No - Mode: Inference - Network: Mobilenet - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: Mobilenet - Device: CPU2 x EPYC 72522 x EPYC 72622 x EPYC 72722 x EPYC 74522 x EPYC 77422 x EPYC 7F322 x EPYC 7F722 x Xeon Gold 6230R2 x Xeon Silver 4210REPYC 7232PEPYC 7302PEPYC 7402PEPYC 7502PXeon Platinum 828048121620SE +/- 0.03, N = 3SE +/- 0.07, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.09, N = 3SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.12, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.06, N = 3SE +/- 0.02, N = 3SE +/- 0.06, N = 3SE +/- 0.11, N = 311.3211.3111.5711.4611.3812.2111.9312.018.4512.1813.6114.2414.5415.18

PlaidML

FP16: No - Mode: Inference - Network: ResNet 50 - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: ResNet 50 - Device: CPU2 x EPYC 72522 x EPYC 72622 x EPYC 72722 x EPYC 74522 x EPYC 77422 x EPYC 7F322 x EPYC 7F722 x Xeon Gold 6230R2 x Xeon Silver 4210REPYC 7232PEPYC 7302PEPYC 7402PEPYC 7502PXeon Platinum 8280246810SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.07, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 34.454.534.514.454.504.824.764.793.655.175.675.775.976.28

PlaidML

FP16: No - Mode: Inference - Network: DenseNet 201 - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: DenseNet 201 - Device: CPU2 x EPYC 72522 x EPYC 72622 x EPYC 72722 x EPYC 74522 x EPYC 77422 x EPYC 7F322 x EPYC 7F722 x Xeon Silver 4210REPYC 7232PEPYC 7302PEPYC 7402PEPYC 7502PXeon Platinum 82800.69081.38162.07242.76323.454SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 32.702.732.712.672.592.992.881.632.623.003.003.072.57

PlaidML

FP16: No - Mode: Inference - Network: Inception V3 - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: Inception V3 - Device: CPU2 x EPYC 72522 x EPYC 72622 x EPYC 72722 x EPYC 74522 x EPYC 77422 x EPYC 7F322 x EPYC 7F722 x Xeon Silver 4210REPYC 7232PEPYC 7302PEPYC 7402PEPYC 7502PXeon Platinum 82803691215SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.13, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.06, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 38.187.978.558.227.788.678.754.687.789.9710.8611.037.35

Phoronix Test Suite v10.8.4