machine learning

AMD Ryzen 9 3900X 12-Core testing with a MSI X570-A PRO (MS-7C37) v3.0 (H.70 BIOS) and NVIDIA GeForce RTX 3060 12GB on Ubuntu 23.10 via the Phoronix Test Suite.

HTML result view exported from: https://openbenchmarking.org/result/2404287-VPA1-DESKTOP70&sro&grs.

machine learningProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDisplay ServerDisplay DriverOpenCLCompilerFile-SystemScreen Resolutionmanticmantic-no-omit-framepointerAMD Ryzen 9 3900X 12-Core @ 3.80GHz (12 Cores / 24 Threads)MSI X570-A PRO (MS-7C37) v3.0 (H.70 BIOS)AMD Starship/Matisse2 x 16GB DDR4-3200MT/s F4-3200C16-16GVK2000GB Seagate ST2000DM006-2DM1 + 2000GB Western Digital WD20EZAZ-00G + 500GB Samsung SSD 860 + 8002GB Seagate ST8000DM004-2CX1 + 1000GB CT1000BX500SSD1 + 512GB TS512GESD310CNVIDIA GeForce RTX 3060 12GBNVIDIA GA104 HD AudioDELL P2314HRealtek RTL8111/8168/8411Ubuntu 23.106.5.0-9-generic (x86_64)X Server 1.21.1.7NVIDIAOpenCL 3.0 CUDA 12.2.146GCC 13.2.0 + CUDA 12.2ext41920x1080NVIDIA GeForce RTX 3060OpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseCompiler Details- mantic: --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,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-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/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-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 - mantic-no-omit-framepointer: --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,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-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-b9QCDx/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-b9QCDx/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-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- Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0x8701013Python 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 Environment Details- mantic-no-omit-framepointer: CXXFLAGS=-fno-omit-frame-pointer QMAKE_CFLAGS=-fno-omit-frame-pointer CFLAGS=-fno-omit-frame-pointer CFLAGS_OVERRIDE=-fno-omit-frame-pointer QMAKE_CXXFLAGS=-fno-omit-frame-pointer FFLAGS=-fno-omit-frame-pointer

machine learningpyhpc: GPU - Numpy - 16384 - Isoneutral Mixingpyhpc: CPU - Numpy - 16384 - Isoneutral Mixingscikit-learn: Treepytorch: NVIDIA CUDA GPU - 16 - Efficientnet_v2_lpyhpc: GPU - Numpy - 262144 - Equation of Statepyperformance: json_loadsscikit-learn: LocalOutlierFactorpytorch: NVIDIA CUDA GPU - 1 - Efficientnet_v2_lpyhpc: CPU - Numpy - 262144 - Equation of Statescikit-learn: Text Vectorizerspyperformance: raytracescikit-learn: Plot Neighborspyperformance: django_templatepyperformance: regex_compilescikit-learn: Sparse Rand Projections / 100 Iterationspytorch: NVIDIA CUDA GPU - 64 - ResNet-152pyperformance: pathlibpyhpc: GPU - Numpy - 262144 - Isoneutral Mixingpyperformance: crypto_pyaespytorch: NVIDIA CUDA GPU - 1 - ResNet-152pytorch: NVIDIA CUDA GPU - 64 - ResNet-50scikit-learn: Isotonic / Perturbed Logarithmscikit-learn: Hist Gradient Boosting Adultpytorch: NVIDIA CUDA GPU - 256 - Efficientnet_v2_lpybench: Total For Average Test Timesscikit-learn: Sample Without Replacementpytorch: NVIDIA CUDA GPU - 512 - ResNet-152scikit-learn: Kernel PCA Solvers / Time vs. N Componentspytorch: NVIDIA CUDA GPU - 64 - Efficientnet_v2_lpyhpc: CPU - Numpy - 4194304 - Isoneutral Mixingpytorch: NVIDIA CUDA GPU - 256 - ResNet-152pytorch: CPU - 32 - ResNet-152pytorch: NVIDIA CUDA GPU - 32 - ResNet-50pyhpc: GPU - Numpy - 4194304 - Isoneutral Mixingpyperformance: gopyperformance: pickle_pure_pythonscikit-learn: Hist Gradient Boosting Categorical Onlypytorch: NVIDIA CUDA GPU - 32 - Efficientnet_v2_lscikit-learn: Covertype Dataset Benchmarkscikit-learn: Sparsifypyhpc: GPU - Numpy - 1048576 - Isoneutral Mixingpytorch: CPU - 256 - ResNet-152scikit-learn: Plot Hierarchicalscikit-learn: Feature Expansionspyperformance: 2to3pyperformance: chaosscikit-learn: Plot OMP vs. LARSpyperformance: nbodyscikit-learn: Hist Gradient Boostingpyhpc: GPU - Numpy - 1048576 - Equation of Statescikit-learn: SGD Regressionpytorch: NVIDIA CUDA GPU - 32 - ResNet-152pytorch: NVIDIA CUDA GPU - 16 - ResNet-152pytorch: NVIDIA CUDA GPU - 512 - ResNet-50pyhpc: GPU - Numpy - 4194304 - Equation of Statepyhpc: CPU - Numpy - 262144 - Isoneutral Mixingscikit-learn: SGDOneClassSVMpyperformance: floatpytorch: CPU - 512 - ResNet-152pytorch: CPU - 512 - Efficientnet_v2_lscikit-learn: SAGApytorch: CPU - 64 - ResNet-50pytorch: CPU - 512 - ResNet-50pytorch: NVIDIA CUDA GPU - 512 - Efficientnet_v2_lpytorch: CPU - 1 - ResNet-50numpy: pytorch: CPU - 256 - Efficientnet_v2_lpytorch: CPU - 64 - Efficientnet_v2_lscikit-learn: GLMscikit-learn: Kernel PCA Solvers / Time vs. N Samplespytorch: CPU - 16 - ResNet-152scikit-learn: 20 Newsgroups / Logistic Regressionscikit-learn: Plot Wardpytorch: CPU - 1 - ResNet-152scikit-learn: Lassopytorch: CPU - 16 - ResNet-50pyperformance: python_startuppyhpc: CPU - Numpy - 1048576 - Equation of Statepytorch: CPU - 64 - ResNet-152pytorch: NVIDIA CUDA GPU - 1 - ResNet-50pytorch: CPU - 32 - ResNet-50pytorch: NVIDIA CUDA GPU - 256 - ResNet-50scikit-learn: Plot Polynomial Kernel Approximationpyhpc: CPU - Numpy - 4194304 - Equation of Statepytorch: CPU - 256 - ResNet-50pytorch: CPU - 32 - Efficientnet_v2_lpytorch: CPU - 16 - Efficientnet_v2_lscikit-learn: MNIST Datasetscikit-learn: Plot Incremental PCApyhpc: CPU - Numpy - 1048576 - Isoneutral Mixingscikit-learn: Hist Gradient Boosting Threadingpytorch: CPU - 1 - Efficientnet_v2_lscikit-learn: Isotonic / Logisticpytorch: NVIDIA CUDA GPU - 16 - ResNet-50scikit-learn: TSNE MNIST Datasetpyhpc: GPU - Numpy - 65536 - Isoneutral Mixingpyhpc: GPU - Numpy - 65536 - Equation of Statepyhpc: CPU - Numpy - 65536 - Isoneutral Mixingpyhpc: CPU - Numpy - 65536 - Equation of Statepyhpc: CPU - Numpy - 16384 - Equation of Statescikit-learn: Isolation Forestpyhpc: GPU - Numpy - 16384 - Equation of Statemanticmantic-no-omit-framepointer0.0090.00948.33838.950.06219.553.46439.350.06160.814262147.75228.5116613.54771.8119.70.13165.173.91201.411788.259103.49737.36774158.26272.3137.24237.882.67071.749.84199.462.66212925918.57937.71376.145127.2820.6319.77211.286131.27722162.891.49976.2109.9840.263106.31574.1573.01203.181.4220.131379.73967.49.875.61868.01824.2424.1337.4332.36426.285.615.62293.59872.5419.8841.51957.82412.72511.84824.287.610.2639.88210.8824.29202.72150.7321.40224.425.635.6365.76331.0060.619110.2157.311470.806200.30236.8650.0330.0150.0320.0150.003289.3710.0030.0080.00852.96936.100.05820.856.75437.290.05863.875274142.45129.5120631.07173.6520.20.12866.672.27205.951828.300105.64736.60790161.46073.7537.88937.242.62672.9110.00202.682.62013126318.86537.16370.694125.4420.6229.91208.391133.09222463.692.58277.1111.2550.260107.52773.3672.24201.141.4110.132382.61166.99.805.65873.82224.4024.2837.2232.54428.615.645.65295.09672.9099.9341.72857.54512.78509.53724.387.640.2629.91211.4624.35203.22150.3761.40524.375.645.6465.87731.0570.618110.3747.321471.834200.17236.7860.0330.0150.0320.0150.003336.3720.002OpenBenchmarking.org

PyHPC Benchmarks

Device: GPU - Backend: Numpy - Project Size: 16384 - Benchmark: Isoneutral Mixing

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: GPU - Backend: Numpy - Project Size: 16384 - Benchmark: Isoneutral Mixingmanticmantic-no-omit-framepointer0.0020.0040.0060.0080.01SE +/- 0.000, N = 3SE +/- 0.000, N = 30.0090.008

PyHPC Benchmarks

Device: CPU - Backend: Numpy - Project Size: 16384 - Benchmark: Isoneutral Mixing

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: CPU - Backend: Numpy - Project Size: 16384 - Benchmark: Isoneutral Mixingmanticmantic-no-omit-framepointer0.0020.0040.0060.0080.01SE +/- 0.000, N = 3SE +/- 0.000, N = 30.0090.008

Scikit-Learn

Benchmark: Tree

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Treemanticmantic-no-omit-framepointer1224364860SE +/- 0.59, N = 4SE +/- 0.48, N = 1548.3452.971. (F9X) gfortran options: -O0

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 16 - Model: Efficientnet_v2_l

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 16 - Model: Efficientnet_v2_lmanticmantic-no-omit-framepointer918273645SE +/- 0.08, N = 3SE +/- 0.02, N = 338.9536.10MIN: 37.12 / MAX: 39.27MIN: 34.25 / MAX: 38.01

PyHPC Benchmarks

Device: GPU - Backend: Numpy - Project Size: 262144 - Benchmark: Equation of State

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: GPU - Backend: Numpy - Project Size: 262144 - Benchmark: Equation of Statemanticmantic-no-omit-framepointer0.0140.0280.0420.0560.07SE +/- 0.001, N = 3SE +/- 0.000, N = 30.0620.058

PyPerformance

Benchmark: json_loads

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: json_loadsmanticmantic-no-omit-framepointer510152025SE +/- 0.06, N = 3SE +/- 0.03, N = 319.520.8

Scikit-Learn

Benchmark: LocalOutlierFactor

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: LocalOutlierFactormanticmantic-no-omit-framepointer1326395265SE +/- 0.18, N = 3SE +/- 0.74, N = 1553.4656.751. (F9X) gfortran options: -O0

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 1 - Model: Efficientnet_v2_l

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 1 - Model: Efficientnet_v2_lmanticmantic-no-omit-framepointer918273645SE +/- 0.47, N = 3SE +/- 0.26, N = 339.3537.29MIN: 36.65 / MAX: 40.42MIN: 35.83 / MAX: 39.17

PyHPC Benchmarks

Device: CPU - Backend: Numpy - Project Size: 262144 - Benchmark: Equation of State

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: CPU - Backend: Numpy - Project Size: 262144 - Benchmark: Equation of Statemanticmantic-no-omit-framepointer0.01370.02740.04110.05480.0685SE +/- 0.001, N = 3SE +/- 0.000, N = 30.0610.058

Scikit-Learn

Benchmark: Text Vectorizers

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Text Vectorizersmanticmantic-no-omit-framepointer1428425670SE +/- 0.19, N = 3SE +/- 0.08, N = 360.8163.881. (F9X) gfortran options: -O0

PyPerformance

Benchmark: raytrace

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: raytracemanticmantic-no-omit-framepointer60120180240300SE +/- 0.33, N = 3SE +/- 0.33, N = 3262274

Scikit-Learn

Benchmark: Plot Neighbors

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Plot Neighborsmanticmantic-no-omit-framepointer306090120150SE +/- 1.34, N = 7SE +/- 0.59, N = 3147.75142.451. (F9X) gfortran options: -O0

PyPerformance

Benchmark: django_template

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: django_templatemanticmantic-no-omit-framepointer714212835SE +/- 0.03, N = 3SE +/- 0.06, N = 328.529.5

PyPerformance

Benchmark: regex_compile

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: regex_compilemanticmantic-no-omit-framepointer306090120150SE +/- 0.00, N = 3SE +/- 0.33, N = 3116120

Scikit-Learn

Benchmark: Sparse Random Projections / 100 Iterations

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Sparse Random Projections / 100 Iterationsmanticmantic-no-omit-framepointer140280420560700SE +/- 3.80, N = 3SE +/- 7.06, N = 4613.55631.071. (F9X) gfortran options: -O0

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 64 - Model: ResNet-152

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 64 - Model: ResNet-152manticmantic-no-omit-framepointer1632486480SE +/- 0.44, N = 3SE +/- 0.66, N = 371.8173.65MIN: 67.31 / MAX: 72.89MIN: 68.88 / MAX: 75.03

PyPerformance

Benchmark: pathlib

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: pathlibmanticmantic-no-omit-framepointer510152025SE +/- 0.00, N = 3SE +/- 0.00, N = 319.720.2

PyHPC Benchmarks

Device: GPU - Backend: Numpy - Project Size: 262144 - Benchmark: Isoneutral Mixing

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: GPU - Backend: Numpy - Project Size: 262144 - Benchmark: Isoneutral Mixingmanticmantic-no-omit-framepointer0.02950.0590.08850.1180.1475SE +/- 0.000, N = 3SE +/- 0.001, N = 30.1310.128

PyPerformance

Benchmark: crypto_pyaes

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: crypto_pyaesmanticmantic-no-omit-framepointer1530456075SE +/- 0.06, N = 3SE +/- 0.00, N = 365.166.6

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 1 - Model: ResNet-152

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 1 - Model: ResNet-152manticmantic-no-omit-framepointer1632486480SE +/- 0.56, N = 3SE +/- 0.96, N = 373.9172.27MIN: 68.9 / MAX: 75.9MIN: 68.86 / MAX: 76.62

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 64 - Model: ResNet-50

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 64 - Model: ResNet-50manticmantic-no-omit-framepointer50100150200250SE +/- 0.58, N = 3SE +/- 1.98, N = 3201.41205.95MIN: 184.02 / MAX: 203.68MIN: 186.96 / MAX: 210.21

Scikit-Learn

Benchmark: Isotonic / Perturbed Logarithm

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Isotonic / Perturbed Logarithmmanticmantic-no-omit-framepointer400800120016002000SE +/- 24.41, N = 3SE +/- 16.46, N = 31788.261828.301. (F9X) gfortran options: -O0

Scikit-Learn

Benchmark: Hist Gradient Boosting Adult

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Hist Gradient Boosting Adultmanticmantic-no-omit-framepointer20406080100SE +/- 0.70, N = 3SE +/- 0.59, N = 3103.50105.651. (F9X) gfortran options: -O0

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 256 - Model: Efficientnet_v2_l

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 256 - Model: Efficientnet_v2_lmanticmantic-no-omit-framepointer918273645SE +/- 0.15, N = 3SE +/- 0.30, N = 1537.3636.60MIN: 35.47 / MAX: 37.85MIN: 33.07 / MAX: 39.53

PyBench

Total For Average Test Times

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyBench 2018-02-16Total For Average Test Timesmanticmantic-no-omit-framepointer2004006008001000SE +/- 1.00, N = 3SE +/- 1.20, N = 3774790

Scikit-Learn

Benchmark: Sample Without Replacement

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Sample Without Replacementmanticmantic-no-omit-framepointer4080120160200SE +/- 0.60, N = 3SE +/- 0.62, N = 3158.26161.461. (F9X) gfortran options: -O0

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 512 - Model: ResNet-152

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 512 - Model: ResNet-152manticmantic-no-omit-framepointer1632486480SE +/- 0.94, N = 3SE +/- 0.50, N = 372.3173.75MIN: 67.38 / MAX: 74.62MIN: 68.91 / MAX: 75.15

Scikit-Learn

Benchmark: Kernel PCA Solvers / Time vs. N Components

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Kernel PCA Solvers / Time vs. N Componentsmanticmantic-no-omit-framepointer918273645SE +/- 0.21, N = 3SE +/- 0.36, N = 337.2437.891. (F9X) gfortran options: -O0

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 64 - Model: Efficientnet_v2_l

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 64 - Model: Efficientnet_v2_lmanticmantic-no-omit-framepointer918273645SE +/- 0.30, N = 9SE +/- 0.31, N = 1537.8837.24MIN: 35.67 / MAX: 39.63MIN: 33.97 / MAX: 39.43

PyHPC Benchmarks

Device: CPU - Backend: Numpy - Project Size: 4194304 - Benchmark: Isoneutral Mixing

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: CPU - Backend: Numpy - Project Size: 4194304 - Benchmark: Isoneutral Mixingmanticmantic-no-omit-framepointer0.60081.20161.80242.40323.004SE +/- 0.010, N = 3SE +/- 0.002, N = 32.6702.626

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 256 - Model: ResNet-152

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 256 - Model: ResNet-152manticmantic-no-omit-framepointer1632486480SE +/- 0.24, N = 3SE +/- 0.83, N = 371.7472.91MIN: 67.87 / MAX: 72.6MIN: 68 / MAX: 75.45

PyTorch

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

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 32 - Model: ResNet-152manticmantic-no-omit-framepointer3691215SE +/- 0.05, N = 3SE +/- 0.09, N = 39.8410.00MIN: 9.6 / MAX: 9.98MIN: 8.09 / MAX: 10.27

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 32 - Model: ResNet-50

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 32 - Model: ResNet-50manticmantic-no-omit-framepointer4080120160200SE +/- 1.06, N = 3SE +/- 2.52, N = 4199.46202.68MIN: 182.77 / MAX: 206.03MIN: 182.69 / MAX: 211.53

PyHPC Benchmarks

Device: GPU - Backend: Numpy - Project Size: 4194304 - Benchmark: Isoneutral Mixing

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: GPU - Backend: Numpy - Project Size: 4194304 - Benchmark: Isoneutral Mixingmanticmantic-no-omit-framepointer0.5991.1981.7972.3962.995SE +/- 0.006, N = 3SE +/- 0.006, N = 32.6622.620

PyPerformance

Benchmark: go

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: gomanticmantic-no-omit-framepointer306090120150SE +/- 0.00, N = 3SE +/- 0.33, N = 3129131

PyPerformance

Benchmark: pickle_pure_python

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: pickle_pure_pythonmanticmantic-no-omit-framepointer60120180240300SE +/- 0.33, N = 3SE +/- 0.58, N = 3259263

Scikit-Learn

Benchmark: Hist Gradient Boosting Categorical Only

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Hist Gradient Boosting Categorical Onlymanticmantic-no-omit-framepointer510152025SE +/- 0.06, N = 3SE +/- 0.12, N = 318.5818.871. (F9X) gfortran options: -O0

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 32 - Model: Efficientnet_v2_l

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 32 - Model: Efficientnet_v2_lmanticmantic-no-omit-framepointer918273645SE +/- 0.24, N = 3SE +/- 0.30, N = 1537.7137.16MIN: 35.52 / MAX: 38.25MIN: 34.12 / MAX: 39.48

Scikit-Learn

Benchmark: Covertype Dataset Benchmark

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Covertype Dataset Benchmarkmanticmantic-no-omit-framepointer80160240320400SE +/- 4.88, N = 3SE +/- 3.40, N = 3376.15370.691. (F9X) gfortran options: -O0

Scikit-Learn

Benchmark: Sparsify

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Sparsifymanticmantic-no-omit-framepointer306090120150SE +/- 1.36, N = 5SE +/- 1.28, N = 5127.28125.441. (F9X) gfortran options: -O0

PyHPC Benchmarks

Device: GPU - Backend: Numpy - Project Size: 1048576 - Benchmark: Isoneutral Mixing

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: GPU - Backend: Numpy - Project Size: 1048576 - Benchmark: Isoneutral Mixingmanticmantic-no-omit-framepointer0.1420.2840.4260.5680.71SE +/- 0.002, N = 3SE +/- 0.007, N = 30.6310.622

PyTorch

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

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: ResNet-152manticmantic-no-omit-framepointer3691215SE +/- 0.07, N = 3SE +/- 0.04, N = 39.779.91MIN: 9.17 / MAX: 10MIN: 9.19 / MAX: 10.05

Scikit-Learn

Benchmark: Plot Hierarchical

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Plot Hierarchicalmanticmantic-no-omit-framepointer50100150200250SE +/- 0.75, N = 3SE +/- 0.42, N = 3211.29208.391. (F9X) gfortran options: -O0

Scikit-Learn

Benchmark: Feature Expansions

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Feature Expansionsmanticmantic-no-omit-framepointer306090120150SE +/- 0.86, N = 3SE +/- 1.22, N = 3131.28133.091. (F9X) gfortran options: -O0

PyPerformance

Benchmark: 2to3

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: 2to3manticmantic-no-omit-framepointer50100150200250SE +/- 0.00, N = 3SE +/- 0.33, N = 3221224

PyPerformance

Benchmark: chaos

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: chaosmanticmantic-no-omit-framepointer1428425670SE +/- 0.03, N = 3SE +/- 0.20, N = 362.863.6

Scikit-Learn

Benchmark: Plot OMP vs. LARS

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Plot OMP vs. LARSmanticmantic-no-omit-framepointer20406080100SE +/- 0.08, N = 3SE +/- 0.44, N = 391.5092.581. (F9X) gfortran options: -O0

PyPerformance

Benchmark: nbody

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: nbodymanticmantic-no-omit-framepointer20406080100SE +/- 0.06, N = 3SE +/- 0.07, N = 376.277.1

Scikit-Learn

Benchmark: Hist Gradient Boosting

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Hist Gradient Boostingmanticmantic-no-omit-framepointer20406080100SE +/- 0.22, N = 3SE +/- 0.25, N = 3109.98111.261. (F9X) gfortran options: -O0

PyHPC Benchmarks

Device: GPU - Backend: Numpy - Project Size: 1048576 - Benchmark: Equation of State

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: GPU - Backend: Numpy - Project Size: 1048576 - Benchmark: Equation of Statemanticmantic-no-omit-framepointer0.05920.11840.17760.23680.296SE +/- 0.002, N = 3SE +/- 0.001, N = 30.2630.260

Scikit-Learn

Benchmark: SGD Regression

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: SGD Regressionmanticmantic-no-omit-framepointer20406080100SE +/- 1.06, N = 6SE +/- 0.49, N = 3106.32107.531. (F9X) gfortran options: -O0

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 32 - Model: ResNet-152

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 32 - Model: ResNet-152manticmantic-no-omit-framepointer1632486480SE +/- 0.96, N = 3SE +/- 0.74, N = 374.1573.36MIN: 68.27 / MAX: 75.61MIN: 68.19 / MAX: 74.63

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 16 - Model: ResNet-152

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 16 - Model: ResNet-152manticmantic-no-omit-framepointer1632486480SE +/- 0.96, N = 3SE +/- 0.20, N = 373.0172.24MIN: 68.06 / MAX: 75.3MIN: 68.36 / MAX: 73.14

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 512 - Model: ResNet-50

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 512 - Model: ResNet-50manticmantic-no-omit-framepointer4080120160200SE +/- 1.69, N = 3SE +/- 0.33, N = 3203.18201.14MIN: 183.76 / MAX: 207.98MIN: 183.61 / MAX: 202.73

PyHPC Benchmarks

Device: GPU - Backend: Numpy - Project Size: 4194304 - Benchmark: Equation of State

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: GPU - Backend: Numpy - Project Size: 4194304 - Benchmark: Equation of Statemanticmantic-no-omit-framepointer0.320.640.961.281.6SE +/- 0.004, N = 3SE +/- 0.001, N = 31.4221.411

PyHPC Benchmarks

Device: CPU - Backend: Numpy - Project Size: 262144 - Benchmark: Isoneutral Mixing

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: CPU - Backend: Numpy - Project Size: 262144 - Benchmark: Isoneutral Mixingmanticmantic-no-omit-framepointer0.02970.05940.08910.11880.1485SE +/- 0.001, N = 3SE +/- 0.000, N = 30.1310.132

Scikit-Learn

Benchmark: SGDOneClassSVM

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: SGDOneClassSVMmanticmantic-no-omit-framepointer80160240320400SE +/- 4.18, N = 3SE +/- 3.48, N = 7379.74382.611. (F9X) gfortran options: -O0

PyPerformance

Benchmark: float

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: floatmanticmantic-no-omit-framepointer1530456075SE +/- 0.03, N = 3SE +/- 0.10, N = 367.466.9

PyTorch

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

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 512 - Model: ResNet-152manticmantic-no-omit-framepointer3691215SE +/- 0.02, N = 3SE +/- 0.07, N = 39.879.80MIN: 9.09 / MAX: 9.96MIN: 9.12 / MAX: 9.98

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_lmanticmantic-no-omit-framepointer1.27132.54263.81395.08526.3565SE +/- 0.01, N = 3SE +/- 0.02, N = 35.615.65MIN: 5.45 / MAX: 5.66MIN: 5.36 / MAX: 5.93

Scikit-Learn

Benchmark: SAGA

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: SAGAmanticmantic-no-omit-framepointer2004006008001000SE +/- 8.69, N = 6SE +/- 5.60, N = 3868.02873.821. (F9X) gfortran options: -O0

PyTorch

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

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 64 - Model: ResNet-50manticmantic-no-omit-framepointer612182430SE +/- 0.04, N = 3SE +/- 0.15, N = 324.2424.40MIN: 23.59 / MAX: 24.49MIN: 21.6 / MAX: 24.8

PyTorch

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

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 512 - Model: ResNet-50manticmantic-no-omit-framepointer612182430SE +/- 0.02, N = 3SE +/- 0.08, N = 324.1324.28MIN: 23.58 / MAX: 24.41MIN: 22.31 / MAX: 24.53

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 512 - Model: Efficientnet_v2_l

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 512 - Model: Efficientnet_v2_lmanticmantic-no-omit-framepointer918273645SE +/- 0.03, N = 3SE +/- 0.33, N = 837.4337.22MIN: 35.81 / MAX: 38.02MIN: 34.99 / MAX: 39.08

PyTorch

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

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: ResNet-50manticmantic-no-omit-framepointer816243240SE +/- 0.11, N = 3SE +/- 0.16, N = 332.3632.54MIN: 31.89 / MAX: 32.7MIN: 31.64 / MAX: 32.94

Numpy Benchmark

OpenBenchmarking.orgScore, More Is BetterNumpy Benchmarkmanticmantic-no-omit-framepointer90180270360450SE +/- 1.20, N = 3SE +/- 0.90, N = 3426.28428.61

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_lmanticmantic-no-omit-framepointer1.2692.5383.8075.0766.345SE +/- 0.02, N = 3SE +/- 0.01, N = 35.615.64MIN: 5.44 / MAX: 5.65MIN: 5.29 / MAX: 5.68

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_lmanticmantic-no-omit-framepointer1.27132.54263.81395.08526.3565SE +/- 0.01, N = 3SE +/- 0.01, N = 35.625.65MIN: 5.35 / MAX: 5.66MIN: 5.45 / MAX: 5.7

Scikit-Learn

Benchmark: GLM

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: GLMmanticmantic-no-omit-framepointer60120180240300SE +/- 1.06, N = 3SE +/- 1.07, N = 3293.60295.101. (F9X) gfortran options: -O0

Scikit-Learn

Benchmark: Kernel PCA Solvers / Time vs. N Samples

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Kernel PCA Solvers / Time vs. N Samplesmanticmantic-no-omit-framepointer1632486480SE +/- 0.05, N = 3SE +/- 0.16, N = 372.5472.911. (F9X) gfortran options: -O0

PyTorch

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

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: ResNet-152manticmantic-no-omit-framepointer3691215SE +/- 0.04, N = 3SE +/- 0.01, N = 39.889.93MIN: 9.31 / MAX: 10.01MIN: 9.39 / MAX: 10.01

Scikit-Learn

Benchmark: 20 Newsgroups / Logistic Regression

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: 20 Newsgroups / Logistic Regressionmanticmantic-no-omit-framepointer1020304050SE +/- 0.19, N = 3SE +/- 0.24, N = 341.5241.731. (F9X) gfortran options: -O0

Scikit-Learn

Benchmark: Plot Ward

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Plot Wardmanticmantic-no-omit-framepointer1326395265SE +/- 0.21, N = 3SE +/- 0.22, N = 357.8257.551. (F9X) gfortran options: -O0

PyTorch

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

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: ResNet-152manticmantic-no-omit-framepointer3691215SE +/- 0.03, N = 3SE +/- 0.04, N = 312.7212.78MIN: 11.99 / MAX: 12.8MIN: 11.9 / MAX: 12.9

Scikit-Learn

Benchmark: Lasso

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Lassomanticmantic-no-omit-framepointer110220330440550SE +/- 3.22, N = 3SE +/- 3.50, N = 3511.85509.541. (F9X) gfortran options: -O0

PyTorch

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

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: ResNet-50manticmantic-no-omit-framepointer612182430SE +/- 0.05, N = 3SE +/- 0.16, N = 324.2824.38MIN: 20.22 / MAX: 24.56MIN: 22.2 / MAX: 24.87

PyPerformance

Benchmark: python_startup

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: python_startupmanticmantic-no-omit-framepointer246810SE +/- 0.01, N = 3SE +/- 0.01, N = 37.617.64

PyHPC Benchmarks

Device: CPU - Backend: Numpy - Project Size: 1048576 - Benchmark: Equation of State

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: CPU - Backend: Numpy - Project Size: 1048576 - Benchmark: Equation of Statemanticmantic-no-omit-framepointer0.05920.11840.17760.23680.296SE +/- 0.002, N = 3SE +/- 0.000, N = 30.2630.262

PyTorch

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

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 64 - Model: ResNet-152manticmantic-no-omit-framepointer3691215SE +/- 0.03, N = 3SE +/- 0.02, N = 39.889.91MIN: 8.8 / MAX: 9.98MIN: 8.69 / MAX: 10.08

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 1 - Model: ResNet-50

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 1 - Model: ResNet-50manticmantic-no-omit-framepointer50100150200250SE +/- 2.67, N = 3SE +/- 1.46, N = 15210.88211.46MIN: 195.21 / MAX: 218.16MIN: 192.13 / MAX: 223.01

PyTorch

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

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 32 - Model: ResNet-50manticmantic-no-omit-framepointer612182430SE +/- 0.10, N = 3SE +/- 0.16, N = 324.2924.35MIN: 22.24 / MAX: 24.66MIN: 23.67 / MAX: 24.87

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 256 - Model: ResNet-50

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 256 - Model: ResNet-50manticmantic-no-omit-framepointer4080120160200SE +/- 1.76, N = 3SE +/- 1.21, N = 3202.72203.22MIN: 183.1 / MAX: 207.93MIN: 185.88 / MAX: 206.71

Scikit-Learn

Benchmark: Plot Polynomial Kernel Approximation

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Plot Polynomial Kernel Approximationmanticmantic-no-omit-framepointer306090120150SE +/- 1.22, N = 3SE +/- 1.20, N = 3150.73150.381. (F9X) gfortran options: -O0

PyHPC Benchmarks

Device: CPU - Backend: Numpy - Project Size: 4194304 - Benchmark: Equation of State

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: CPU - Backend: Numpy - Project Size: 4194304 - Benchmark: Equation of Statemanticmantic-no-omit-framepointer0.31610.63220.94831.26441.5805SE +/- 0.003, N = 3SE +/- 0.004, N = 31.4021.405

PyTorch

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

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: ResNet-50manticmantic-no-omit-framepointer612182430SE +/- 0.03, N = 3SE +/- 0.11, N = 324.4224.37MIN: 20.15 / MAX: 24.74MIN: 23.76 / MAX: 24.81

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_lmanticmantic-no-omit-framepointer1.2692.5383.8075.0766.345SE +/- 0.01, N = 3SE +/- 0.01, N = 35.635.64MIN: 5.31 / MAX: 5.68MIN: 5.52 / MAX: 5.69

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_lmanticmantic-no-omit-framepointer1.2692.5383.8075.0766.345SE +/- 0.02, N = 3SE +/- 0.01, N = 35.635.64MIN: 5.39 / MAX: 5.71MIN: 5.45 / MAX: 5.68

Scikit-Learn

Benchmark: MNIST Dataset

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: MNIST Datasetmanticmantic-no-omit-framepointer1530456075SE +/- 0.82, N = 4SE +/- 0.47, N = 365.7665.881. (F9X) gfortran options: -O0

Scikit-Learn

Benchmark: Plot Incremental PCA

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Plot Incremental PCAmanticmantic-no-omit-framepointer714212835SE +/- 0.03, N = 3SE +/- 0.07, N = 331.0131.061. (F9X) gfortran options: -O0

PyHPC Benchmarks

Device: CPU - Backend: Numpy - Project Size: 1048576 - Benchmark: Isoneutral Mixing

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: CPU - Backend: Numpy - Project Size: 1048576 - Benchmark: Isoneutral Mixingmanticmantic-no-omit-framepointer0.13930.27860.41790.55720.6965SE +/- 0.001, N = 3SE +/- 0.000, N = 30.6190.618

Scikit-Learn

Benchmark: Hist Gradient Boosting Threading

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Hist Gradient Boosting Threadingmanticmantic-no-omit-framepointer20406080100SE +/- 0.13, N = 3SE +/- 0.15, N = 3110.22110.371. (F9X) gfortran options: -O0

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_lmanticmantic-no-omit-framepointer246810SE +/- 0.00, N = 3SE +/- 0.02, N = 37.317.32MIN: 7.16 / MAX: 7.34MIN: 7.23 / MAX: 7.38

Scikit-Learn

Benchmark: Isotonic / Logistic

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Isotonic / Logisticmanticmantic-no-omit-framepointer30060090012001500SE +/- 12.29, N = 3SE +/- 14.46, N = 31470.811471.831. (F9X) gfortran options: -O0

PyTorch

Device: NVIDIA CUDA GPU - Batch Size: 16 - Model: ResNet-50

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: NVIDIA CUDA GPU - Batch Size: 16 - Model: ResNet-50manticmantic-no-omit-framepointer4080120160200SE +/- 0.25, N = 3SE +/- 0.96, N = 3200.30200.17MIN: 182.88 / MAX: 202.36MIN: 183.43 / MAX: 203.55

Scikit-Learn

Benchmark: TSNE MNIST Dataset

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: TSNE MNIST Datasetmanticmantic-no-omit-framepointer50100150200250SE +/- 0.44, N = 3SE +/- 0.54, N = 3236.87236.791. (F9X) gfortran options: -O0

PyHPC Benchmarks

Device: GPU - Backend: Numpy - Project Size: 65536 - Benchmark: Isoneutral Mixing

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: GPU - Backend: Numpy - Project Size: 65536 - Benchmark: Isoneutral Mixingmanticmantic-no-omit-framepointer0.00740.01480.02220.02960.037SE +/- 0.000, N = 3SE +/- 0.000, N = 30.0330.033

PyHPC Benchmarks

Device: GPU - Backend: Numpy - Project Size: 65536 - Benchmark: Equation of State

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: GPU - Backend: Numpy - Project Size: 65536 - Benchmark: Equation of Statemanticmantic-no-omit-framepointer0.00340.00680.01020.01360.017SE +/- 0.000, N = 3SE +/- 0.000, N = 30.0150.015

PyHPC Benchmarks

Device: CPU - Backend: Numpy - Project Size: 65536 - Benchmark: Isoneutral Mixing

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: CPU - Backend: Numpy - Project Size: 65536 - Benchmark: Isoneutral Mixingmanticmantic-no-omit-framepointer0.00720.01440.02160.02880.036SE +/- 0.000, N = 3SE +/- 0.000, N = 30.0320.032

PyHPC Benchmarks

Device: CPU - Backend: Numpy - Project Size: 65536 - Benchmark: Equation of State

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: CPU - Backend: Numpy - Project Size: 65536 - Benchmark: Equation of Statemanticmantic-no-omit-framepointer0.00340.00680.01020.01360.017SE +/- 0.000, N = 3SE +/- 0.000, N = 30.0150.015

PyHPC Benchmarks

Device: CPU - Backend: Numpy - Project Size: 16384 - Benchmark: Equation of State

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: CPU - Backend: Numpy - Project Size: 16384 - Benchmark: Equation of Statemanticmantic-no-omit-framepointer0.00070.00140.00210.00280.0035SE +/- 0.000, N = 3SE +/- 0.000, N = 30.0030.003

Scikit-Learn

Benchmark: Isolation Forest

OpenBenchmarking.orgSeconds, Fewer Is BetterScikit-Learn 1.2.2Benchmark: Isolation Forestmanticmantic-no-omit-framepointer70140210280350SE +/- 1.30, N = 3SE +/- 51.04, N = 9289.37336.371. (F9X) gfortran options: -O0

PyHPC Benchmarks

Device: GPU - Backend: Numpy - Project Size: 16384 - Benchmark: Equation of State

OpenBenchmarking.orgSeconds, Fewer Is BetterPyHPC Benchmarks 3.0Device: GPU - Backend: Numpy - Project Size: 16384 - Benchmark: Equation of Statemanticmantic-no-omit-framepointer0.00070.00140.00210.00280.0035SE +/- 0.000, N = 3SE +/- 0.000, N = 150.0030.002


Phoronix Test Suite v10.8.5