9684x-march

Tests for a future article. 2 x AMD EPYC 9684X 96-Core testing with a AMD Titanite_4G (RTI1007B BIOS) and ASPEED on Ubuntu 23.10 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2403270-NE-9684XMARC10
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
Show Result Confidence Charts
Allow Limiting Results To Certain Suite(s)

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
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
Prefer Vertical Bar Graphs

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Toggle/Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
PRE
March 27
  2 Hours, 34 Minutes
a
March 27
  8 Hours, 3 Minutes
Invert Behavior (Only Show Selected Data)
  5 Hours, 18 Minutes
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):


9684x-march OpenBenchmarking.orgPhoronix Test Suite2 x AMD EPYC 9684X 96-Core @ 2.55GHz (192 Cores / 384 Threads)AMD Titanite_4G (RTI1007B BIOS)AMD Device 14a41520GB3201GB Micron_7450_MTFDKCB3T2TFS + 257GB Flash DriveASPEEDBroadcom NetXtreme BCM5720 PCIeUbuntu 23.106.5.0-25-generic (x86_64)GCC 13.2.0ext4640x480ProcessorMotherboardChipsetMemoryDiskGraphicsNetworkOSKernelCompilerFile-SystemScreen Resolution9684x-march BenchmarksSystem Logs- Transparent Huge Pages: madvise- --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 - Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa10113e - Python 3.11.6- gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + 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 Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

PRE vs. a ComparisonPhoronix Test SuiteBaseline+2.3%+2.3%+4.6%+4.6%+6.9%+6.9%+9.2%+9.2%7.1%6.1%4.9%4%3.2%2.9%2.9%2.8%2.5%2.2%CPU - 32 - ResNet-509.3%CPU - 32 - ResNet-152CPU - 1 - ResNet-152CPU - 32 - GoogLeNet5%CPU - 1 - GoogLeNetCPU - 16 - ResNet-50CPU - 1 - ResNet-503.8%CPU - 64 - ResNet-1523.4%CPU - 32 - ResNet-50CPU - 256 - AlexNet3%CPU - 32 - AlexNetCPU - 16 - ResNet-50CPU - 512 - ResNet-50Read While Writing2.7%CPU - 1 - Efficientnet_v2_lCPU - 64 - ResNet-502.4%CPU - 16 - AlexNetCPU - 64 - AlexNet2.1%CPU - 256 - ResNet-502.1%TensorFlowPyTorchPyTorchTensorFlowTensorFlowTensorFlowTensorFlowPyTorchPyTorchTensorFlowTensorFlowPyTorchPyTorchRocksDBPyTorchPyTorchTensorFlowTensorFlowPyTorchPREa

9684x-march pytorch: 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_ltensorflow: CPU - 1 - AlexNettensorflow: CPU - 16 - AlexNettensorflow: CPU - 32 - AlexNettensorflow: CPU - 64 - AlexNettensorflow: CPU - 1 - GoogLeNettensorflow: CPU - 1 - ResNet-50tensorflow: CPU - 256 - AlexNettensorflow: CPU - 512 - AlexNettensorflow: CPU - 16 - GoogLeNettensorflow: CPU - 16 - ResNet-50tensorflow: CPU - 32 - GoogLeNettensorflow: CPU - 32 - ResNet-50tensorflow: CPU - 64 - GoogLeNettensorflow: CPU - 64 - ResNet-50tensorflow: CPU - 256 - GoogLeNettensorflow: CPU - 256 - ResNet-50tensorflow: CPU - 512 - GoogLeNettensorflow: CPU - 512 - ResNet-50rocksdb: Overwriterocksdb: Rand Readrocksdb: Update Randrocksdb: Read While Writingrocksdb: Read Rand Write Randbrl-cad: VGR Performance Metricbuild-mesa: Time To Compileblender: BMW27 - CPU-Onlyblender: Junkshop - CPU-Onlyblender: Classroom - CPU-Onlyblender: Fishy Cat - CPU-Onlyblender: Barbershop - CPU-Onlyblender: Pabellon Barcelona - CPU-OnlyPREa23.069.9720.9320.1921.598.9321.208.7220.439.218.929.476.292.332.332.322.292.3121.16242.29424.06765.5512.584.051652.231980.51112.6439.68185.1665.88275.3487.72400.03119.83493.31140.594210491105306233421266271303633619142595661214.667.5511.418.039.9667.3822.9923.2010.5821.5320.8421.089.0120.779.3421.018.919.099.336.452.332.312.312.332.3320.78247.55436.25749.4613.203.91604.522010.56114.2641.26176.3660.25273.6888.93399.46118.88484.02140.494216161108892776425687264066623643263592756414.7567.5511.4418.089.8567.6623.1OpenBenchmarking.org

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 1 - Model: ResNet-50PREa612182430SE +/- 0.20, N = 1523.0623.20MIN: 12.95 / MAX: 24.52MIN: 12.21 / MAX: 25.13

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 1 - Model: ResNet-152PREa3691215SE +/- 0.10, N = 159.9710.58MIN: 4.85 / MAX: 10.69MIN: 4.55 / MAX: 11.67

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 16 - Model: ResNet-50PREa510152025SE +/- 0.16, N = 320.9321.53MIN: 12.91 / MAX: 21.51MIN: 12.64 / MAX: 22.28

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 32 - Model: ResNet-50PREa510152025SE +/- 0.16, N = 1520.1920.84MIN: 11.95 / MAX: 21.04MIN: 11.24 / MAX: 22.33

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 64 - Model: ResNet-50PREa510152025SE +/- 0.23, N = 321.5921.08MIN: 14.02 / MAX: 22.21MIN: 13.2 / MAX: 22.07

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 16 - Model: ResNet-152PREa3691215SE +/- 0.09, N = 38.939.01MIN: 8.8 / MAX: 9.04MIN: 4.81 / MAX: 9.31

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 256 - Model: ResNet-50PREa510152025SE +/- 0.10, N = 321.2020.77MIN: 12.68 / MAX: 21.88MIN: 12.97 / MAX: 21.67

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 32 - Model: ResNet-152PREa3691215SE +/- 0.08, N = 38.729.34MIN: 5.23 / MAX: 9.06MIN: 4.74 / MAX: 9.74

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 512 - Model: ResNet-50PREa510152025SE +/- 0.14, N = 1520.4321.01MIN: 13.46 / MAX: 21.1MIN: 11.92 / MAX: 22.65

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 64 - Model: ResNet-152PREa3691215SE +/- 0.09, N = 129.218.91MIN: 4.8 / MAX: 9.43MIN: 4.5 / MAX: 9.7

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 256 - Model: ResNet-152PREa3691215SE +/- 0.10, N = 128.929.09MIN: 5.04 / MAX: 9.16MIN: 4.84 / MAX: 10.03

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 512 - Model: ResNet-152PREa3691215SE +/- 0.10, N = 39.479.33MIN: 5.17 / MAX: 9.87MIN: 4.69 / MAX: 9.66

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 1 - Model: Efficientnet_v2_lPREa246810SE +/- 0.09, N = 36.296.45MIN: 3.09 / MAX: 6.44MIN: 3.05 / MAX: 6.85

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 16 - Model: Efficientnet_v2_lPREa0.52431.04861.57292.09722.6215SE +/- 0.01, N = 32.332.33MIN: 1.76 / MAX: 2.72MIN: 1.77 / MAX: 2.9

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 32 - Model: Efficientnet_v2_lPREa0.52431.04861.57292.09722.6215SE +/- 0.01, N = 32.332.31MIN: 1.78 / MAX: 2.8MIN: 1.88 / MAX: 2.74

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 64 - Model: Efficientnet_v2_lPREa0.5221.0441.5662.0882.61SE +/- 0.01, N = 32.322.31MIN: 1.9 / MAX: 2.75MIN: 1.53 / MAX: 2.83

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 256 - Model: Efficientnet_v2_lPREa0.52431.04861.57292.09722.6215SE +/- 0.01, N = 32.292.33MIN: 1.79 / MAX: 2.72MIN: 1.59 / MAX: 2.78

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 512 - Model: Efficientnet_v2_lPREa0.52431.04861.57292.09722.6215SE +/- 0.01, N = 32.312.33MIN: 1.7 / MAX: 2.84MIN: 1.58 / MAX: 2.83

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 1 - Model: AlexNetPREa510152025SE +/- 0.16, N = 1521.1620.78

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 16 - Model: AlexNetPREa50100150200250SE +/- 2.30, N = 15242.29247.55

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 32 - Model: AlexNetPREa90180270360450SE +/- 6.62, N = 15424.06436.25

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: AlexNetPREa170340510680850SE +/- 5.39, N = 15765.55749.46

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 1 - Model: GoogLeNetPREa3691215SE +/- 0.14, N = 1512.5813.20

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 1 - Model: ResNet-50PREa0.91131.82262.73393.64524.55654.053.90

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: AlexNetPREa4008001200160020001652.231604.52

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: AlexNetPREa4008001200160020001980.512010.56

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 16 - Model: GoogLeNetPREa306090120150112.64114.26

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 16 - Model: ResNet-50PREa91827364539.6841.26

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 32 - Model: GoogLeNetPREa4080120160200185.16176.36

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 32 - Model: ResNet-50aPRE153045607560.2565.88

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: GoogLeNetaPRE60120180240300273.68275.34

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: ResNet-50aPRE2040608010088.9387.72

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: GoogLeNetaPRE90180270360450399.46400.03

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: ResNet-50aPRE306090120150118.88119.83

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: GoogLeNetaPRE110220330440550484.02493.31

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: ResNet-50aPRE306090120150140.49140.59

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: OverwriteaPRE90K180K270K360K450K4216164210491. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Random ReadaPRE200M400M600M800M1000M110889277611053062331. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Update RandomaPRE90K180K270K360K450K4256874212661. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Read While WritingaPRE6M12M18M24M30M26406662271303631. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Read Random Write RandomaPRE800K1600K2400K3200K4000K364326336191421. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

BRL-CAD

BRL-CAD is a cross-platform, open-source solid modeling system with built-in benchmark mode. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgVGR Performance Metric, More Is BetterBRL-CAD 7.38.2VGR Performance MetricaPRE1.3M2.6M3.9M5.2M6.5M592756459566121. (CXX) g++ options: -std=c++17 -pipe -fvisibility=hidden -fno-strict-aliasing -fno-common -fexceptions -ftemplate-depth-128 -m64 -ggdb3 -O3 -fipa-pta -fstrength-reduce -finline-functions -flto -ltcl8.6 -lnetpbm -lregex_brl -lz_brl -lassimp -ldl -lm -ltk8.6

Timed Mesa Compilation

This test profile times how long it takes to compile Mesa with Meson/Ninja. For minimizing build dependencies and avoid versioning conflicts, test this is just the core Mesa build without LLVM or the extra Gallium3D/Mesa drivers enabled. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Mesa Compilation 24.0Time To CompilePREa48121620SE +/- 0.04, N = 314.6614.76

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: BMW27 - Compute: CPU-OnlyaPRE2468107.557.55

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Junkshop - Compute: CPU-OnlyaPRE369121511.4411.40

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Classroom - Compute: CPU-OnlyaPRE4812162018.0818.03

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Fishy Cat - Compute: CPU-OnlyaPRE36912159.859.96

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Barbershop - Compute: CPU-OnlyaPRE153045607567.6667.38

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Pabellon Barcelona - Compute: CPU-OnlyaPRE61218243023.1022.99