pytorch bench

AMD Ryzen Threadripper 3990X 64-Core testing with a Gigabyte TRX40 AORUS PRO WIFI (F6 BIOS) and AMD Radeon RX 5700 8GB on Ubuntu 23.10 via the Phoronix Test Suite.

HTML result view exported from: https://openbenchmarking.org/result/2311179-PTS-PYTORCHB94.

pytorch benchProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerOpenGLCompilerFile-SystemScreen ResolutionabcAMD Ryzen Threadripper 3990X 64-Core @ 2.90GHz (64 Cores / 128 Threads)Gigabyte TRX40 AORUS PRO WIFI (F6 BIOS)AMD Starship/Matisse128GBSamsung SSD 970 EVO Plus 500GBAMD Radeon RX 5700 8GB (1750/875MHz)AMD Navi 10 HDMI AudioDELL P2415QIntel I211 + Intel Wi-Fi 6 AX200Ubuntu 23.106.5.0-10-generic (x86_64)GNOME Shell 45.0X Server + Wayland4.6 Mesa 23.2.1-1ubuntu3 (LLVM 15.0.7 DRM 3.54)GCC 13.2.0ext43840x2160OpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseProcessor Details- Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0x830107aPython Details- Python 3.11.6Security Details- gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Mitigation of untrained return thunk; SMT enabled with STIBP protection + 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 Retpolines IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

pytorch benchpytorch: CPU - 1 - ResNet-50pytorch: CPU - 1 - ResNet-152pytorch: CPU - 16 - ResNet-50pytorch: CPU - 32 - ResNet-50pytorch: CPU - 64 - ResNet-50pytorch: CPU - 16 - ResNet-152pytorch: CPU - 256 - ResNet-50pytorch: CPU - 32 - ResNet-152pytorch: CPU - 512 - ResNet-50pytorch: CPU - 64 - ResNet-152pytorch: CPU - 256 - ResNet-152pytorch: CPU - 512 - ResNet-152pytorch: CPU - 1 - Efficientnet_v2_lpytorch: CPU - 16 - Efficientnet_v2_lpytorch: CPU - 32 - Efficientnet_v2_lpytorch: CPU - 64 - Efficientnet_v2_lpytorch: CPU - 256 - Efficientnet_v2_lpytorch: CPU - 512 - Efficientnet_v2_labc20.888.2216.8817.1115.966.6916.666.7816.596.847.016.894.363.073.012.962.932.9218.637.8016.0316.0915.996.8116.836.7616.066.826.526.664.172.902.992.922.862.9119.188.0315.5716.5216.356.7616.276.6915.646.676.656.514.162.932.962.963.022.99OpenBenchmarking.org

PyTorch

Device: CPU - Batch Size: 1 - Model: ResNet-50

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: ResNet-50abc510152025SE +/- 0.27, N = 320.8818.6319.18MIN: 19.27 / MAX: 22.47MIN: 17.58 / MAX: 19.74MIN: 18.22 / MAX: 20.52

PyTorch

Device: CPU - Batch Size: 1 - Model: ResNet-152

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: ResNet-152abc246810SE +/- 0.10, N = 38.227.808.03MIN: 7.88 / MAX: 8.56MIN: 7.62 / MAX: 8.01MIN: 7.82 / MAX: 8.21

PyTorch

Device: CPU - Batch Size: 16 - Model: ResNet-50

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: ResNet-50abc48121620SE +/- 0.19, N = 316.8816.0315.57MIN: 15.77 / MAX: 18.22MIN: 14.93 / MAX: 16.68MIN: 14.78 / MAX: 16.29

PyTorch

Device: CPU - Batch Size: 32 - Model: ResNet-50

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 32 - Model: ResNet-50abc48121620SE +/- 0.20, N = 317.1116.0916.52MIN: 15.92 / MAX: 18.06MIN: 15.44 / MAX: 16.98MIN: 15.52 / MAX: 17.2

PyTorch

Device: CPU - Batch Size: 64 - Model: ResNet-50

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 64 - Model: ResNet-50abc48121620SE +/- 0.13, N = 315.9615.9916.35MIN: 14.8 / MAX: 17.08MIN: 15.31 / MAX: 16.83MIN: 15.49 / MAX: 17.11

PyTorch

Device: CPU - Batch Size: 16 - Model: ResNet-152

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: ResNet-152abc246810SE +/- 0.06, N = 36.696.816.76MIN: 6.41 / MAX: 6.91MIN: 6.67 / MAX: 7MIN: 6.34 / MAX: 6.95

PyTorch

Device: CPU - Batch Size: 256 - Model: ResNet-50

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: ResNet-50abc48121620SE +/- 0.21, N = 316.6616.8316.27MIN: 15.65 / MAX: 17.93MIN: 15.98 / MAX: 17.49MIN: 15.45 / MAX: 16.87

PyTorch

Device: CPU - Batch Size: 32 - Model: ResNet-152

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 32 - Model: ResNet-152abc246810SE +/- 0.06, N = 36.786.766.69MIN: 6.21 / MAX: 7.03MIN: 6.63 / MAX: 6.89MIN: 6.25 / MAX: 6.86

PyTorch

Device: CPU - Batch Size: 512 - Model: ResNet-50

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 512 - Model: ResNet-50abc48121620SE +/- 0.14, N = 816.5916.0615.64MIN: 15.41 / MAX: 17.89MIN: 14.68 / MAX: 16.72MIN: 15.06 / MAX: 16.39

PyTorch

Device: CPU - Batch Size: 64 - Model: ResNet-152

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 64 - Model: ResNet-152abc246810SE +/- 0.08, N = 36.846.826.67MIN: 6.51 / MAX: 7.13MIN: 6.66 / MAX: 6.95MIN: 6.48 / MAX: 6.83

PyTorch

Device: CPU - Batch Size: 256 - Model: ResNet-152

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: ResNet-152abc246810SE +/- 0.05, N = 37.016.526.65MIN: 6.51 / MAX: 7.22MIN: 6.32 / MAX: 6.71MIN: 6.45 / MAX: 6.82

PyTorch

Device: CPU - Batch Size: 512 - Model: ResNet-152

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 512 - Model: ResNet-152abc246810SE +/- 0.01, N = 36.896.666.51MIN: 6.48 / MAX: 7.08MIN: 6.29 / MAX: 6.85MIN: 6.34 / MAX: 6.69

PyTorch

Device: CPU - Batch Size: 1 - Model: Efficientnet_v2_l

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: Efficientnet_v2_labc0.9811.9622.9433.9244.905SE +/- 0.04, N = 34.364.174.16MIN: 4.07 / MAX: 4.56MIN: 3.98 / MAX: 4.35MIN: 3.9 / MAX: 4.35

PyTorch

Device: CPU - Batch Size: 16 - Model: Efficientnet_v2_l

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: Efficientnet_v2_labc0.69081.38162.07242.76323.454SE +/- 0.03, N = 33.072.902.93MIN: 2.85 / MAX: 3.2MIN: 2.73 / MAX: 3.05MIN: 2.82 / MAX: 3.07

PyTorch

Device: CPU - Batch Size: 32 - Model: Efficientnet_v2_l

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 32 - Model: Efficientnet_v2_labc0.67731.35462.03192.70923.3865SE +/- 0.04, N = 43.012.992.96MIN: 2.81 / MAX: 3.25MIN: 2.79 / MAX: 3.27MIN: 2.83 / MAX: 3.07

PyTorch

Device: CPU - Batch Size: 64 - Model: Efficientnet_v2_l

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 64 - Model: Efficientnet_v2_labc0.6661.3321.9982.6643.33SE +/- 0.00, N = 32.962.922.96MIN: 2.81 / MAX: 3.09MIN: 2.76 / MAX: 3.04MIN: 2.82 / MAX: 3.09

PyTorch

Device: CPU - Batch Size: 256 - Model: Efficientnet_v2_l

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: Efficientnet_v2_labc0.67951.3592.03852.7183.3975SE +/- 0.02, N = 32.932.863.02MIN: 2.74 / MAX: 3.09MIN: 2.76 / MAX: 2.99MIN: 2.9 / MAX: 3.14

PyTorch

Device: CPU - Batch Size: 512 - Model: Efficientnet_v2_l

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 512 - Model: Efficientnet_v2_labc0.67281.34562.01842.69123.364SE +/- 0.02, N = 32.922.912.99MIN: 2.78 / MAX: 3.12MIN: 2.77 / MAX: 3.03MIN: 2.84 / MAX: 3.16


Phoronix Test Suite v10.8.4