eps

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 2312240-NE-EPS17737430
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

Limit displaying results to tests within:

CPU Massive 3 Tests
Creator Workloads 2 Tests
HPC - High Performance Computing 3 Tests
Java Tests 2 Tests
Machine Learning 3 Tests
Python Tests 3 Tests

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Geometric Means Per-Suite/Category
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
Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
a
December 24 2023
  1 Day, 26 Minutes
b
December 25 2023
  7 Hours, 39 Minutes
Invert Hiding All Results Option
  16 Hours, 3 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):


epsOpenBenchmarking.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_MTFDKCB3T2TFSASPEEDBroadcom NetXtreme BCM5720 PCIeUbuntu 23.106.5.0-13-generic (x86_64)GCC 13.2.0ext4800x600ProcessorMotherboardChipsetMemoryDiskGraphicsNetworkOSKernelCompilerFile-SystemScreen ResolutionEps PerformanceSystem 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 - OpenJDK Runtime Environment (build 11.0.21+9-post-Ubuntu-0ubuntu123.10)- 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

a vs. b ComparisonPhoronix Test SuiteBaseline+7.7%+7.7%+15.4%+15.4%+23.1%+23.1%+30.8%+30.8%7.7%5.3%4.3%4.3%2.7%2.1%13.1%2.3%3.1%9.6%2%2.4%6.7%7.4%2.5%2.7%6.6%10.5%9.6%4.1%9.7%2.7%11.7%3.4%30.8%11.9%4.6%CPU - 256 - ResNet-152CPU - 1 - Efficientnet_v2_lPreset 13 - Bosphorus 4KPreset 12 - Bosphorus 4KQ.1.C.E.53.7%CPU - 256 - ResNet-503.3%CPU - 1 - ResNet-152BLAS50 - Q2150 - Q1915.6%50 - Q1850 - Q165.3%50 - Q1550 - Q132.2%50 - Q1250 - Q1150 - Q074%50 - Q054.6%50 - Q043.9%50 - Q0313.4%50 - Q017.2%10 - Q1910 - Q1810 - Q1510 - Q1410 - Q137.6%10 - Q113.4%10 - Q1010 - Q092.8%10 - Q0810 - Q0610 - Q0514.7%10 - Q0410 - Q032.3%10 - Q011 - Q225.9%1 - Q198.5%1 - Q181 - Q171 - Q169.9%1 - Q153.4%1 - Q147.1%1 - Q139.6%1 - Q124.2%1 - Q111 - Q093.3%1 - Q071 - Q061 - Q051 - Q041 - Q012.9%PyTorchPyTorchSVT-AV1SVT-AV1WebP2 Image EncodePyTorchPyTorchLeelaChessZeroApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-HApache Spark TPC-Hab

epspytorch: CPU - 256 - ResNet-152pytorch: CPU - 1 - Efficientnet_v2_lsvt-av1: Preset 13 - Bosphorus 4Ksvt-av1: Preset 12 - Bosphorus 4Kwebp2: Quality 100, Compression Effort 5pytorch: CPU - 256 - ResNet-50pytorch: CPU - 1 - ResNet-152pytorch: CPU - 1 - ResNet-50pytorch: CPU - 16 - ResNet-50spark-tpch: 1 - Geometric Mean Of All Queriespytorch: CPU - 256 - Efficientnet_v2_lwebp2: Defaultsvt-av1: Preset 8 - Bosphorus 1080pdeepsparse: ResNet-50, Baseline - Synchronous Single-Streamdeepsparse: ResNet-50, Baseline - Synchronous Single-Streamwebp2: Quality 75, Compression Effort 7xmrig: CryptoNight-Femto UPX2 - 1Mpytorch: CPU - 32 - ResNet-152pytorch: CPU - 16 - Efficientnet_v2_ljava-scimark2: Sparse Matrix Multiplyxmrig: CryptoNight-Heavy - 1Mspark-tpch: 10 - Geometric Mean Of All Queriesjava-scimark2: Dense LU Matrix Factorizationsvt-av1: Preset 4 - Bosphorus 1080pdeepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Streamsvt-av1: Preset 13 - Bosphorus 1080psvt-av1: Preset 4 - Bosphorus 4Ksvt-av1: Preset 8 - Bosphorus 4Kdeepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Streampytorch: CPU - 16 - ResNet-152deepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Streampytorch: CPU - 32 - ResNet-50xmrig: GhostRider - 1Mdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Synchronous Single-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Synchronous Single-Streamdeepsparse: ResNet-50, Sparse INT8 - Asynchronous Multi-Streamxmrig: Wownero - 1Mdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamdeepsparse: ResNet-50, Sparse INT8 - Asynchronous Multi-Streamdeepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Streamsvt-av1: Preset 12 - Bosphorus 1080pdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Streamxmrig: Monero - 1Mdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Synchronous Single-Streamdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Synchronous Single-Streamjava-scimark2: Compositedeepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamjava-scimark2: Fast Fourier Transformdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Streamopenssl: SHA512deepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamdeepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Asynchronous Multi-Streamdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Synchronous Single-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Synchronous Single-Streamdeepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Asynchronous Multi-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: BERT-Large, NLP Question Answering - Synchronous Single-Streamdeepsparse: BERT-Large, NLP Question Answering - Synchronous Single-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamopenssl: SHA256xmrig: KawPow - 1Mspark-tpch: 50 - Geometric Mean Of All Queriesdeepsparse: ResNet-50, Baseline - Asynchronous Multi-Streamopenssl: RSA4096deepsparse: ResNet-50, Baseline - Asynchronous Multi-Streamdeepsparse: ResNet-50, Sparse INT8 - Synchronous Single-Streamdeepsparse: ResNet-50, Sparse INT8 - Synchronous Single-Streamjava-scimark2: Monte Carloopenssl: RSA4096deepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Synchronous Single-Streamdeepsparse: CV Detection, YOLOv5s COCO - Synchronous Single-Streamdeepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Synchronous Single-Streamdeepsparse: CV Detection, YOLOv5s COCO - Synchronous Single-Streamjava-scimark2: Jacobi Successive Over-Relaxationdeepsparse: BERT-Large, NLP Question Answering - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamdeepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Synchronous Single-Streamdeepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Synchronous Single-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Streamdeepsparse: BERT-Large, NLP Question Answering - Asynchronous Multi-Streampytorch: CPU - 32 - Efficientnet_v2_lwebp2: Quality 100, Lossless Compressionwebp2: Quality 95, Compression Effort 7spark-tpch: 50 - Q22spark-tpch: 50 - Q21spark-tpch: 50 - Q20spark-tpch: 50 - Q19spark-tpch: 50 - Q18spark-tpch: 50 - Q17spark-tpch: 50 - Q16spark-tpch: 50 - Q15spark-tpch: 50 - Q14spark-tpch: 50 - Q13spark-tpch: 50 - Q12spark-tpch: 50 - Q11spark-tpch: 50 - Q10spark-tpch: 50 - Q09spark-tpch: 50 - Q08spark-tpch: 50 - Q07spark-tpch: 50 - Q06spark-tpch: 50 - Q05spark-tpch: 50 - Q04spark-tpch: 50 - Q03spark-tpch: 50 - Q02spark-tpch: 50 - Q01spark-tpch: 10 - Q22spark-tpch: 10 - Q21spark-tpch: 10 - Q20spark-tpch: 10 - Q19spark-tpch: 10 - Q18spark-tpch: 10 - Q17spark-tpch: 10 - Q16spark-tpch: 10 - Q15spark-tpch: 10 - Q14spark-tpch: 10 - Q13spark-tpch: 10 - Q12spark-tpch: 10 - Q11spark-tpch: 10 - Q10spark-tpch: 10 - Q09spark-tpch: 10 - Q08spark-tpch: 10 - Q07spark-tpch: 10 - Q06spark-tpch: 10 - Q05spark-tpch: 10 - Q04spark-tpch: 10 - Q03spark-tpch: 10 - Q02spark-tpch: 10 - Q01spark-tpch: 1 - Q22spark-tpch: 1 - Q21spark-tpch: 1 - Q20spark-tpch: 1 - Q19spark-tpch: 1 - Q18spark-tpch: 1 - Q17spark-tpch: 1 - Q16spark-tpch: 1 - Q15spark-tpch: 1 - Q14spark-tpch: 1 - Q13spark-tpch: 1 - Q12spark-tpch: 1 - Q11spark-tpch: 1 - Q10spark-tpch: 1 - Q09spark-tpch: 1 - Q08spark-tpch: 1 - Q07spark-tpch: 1 - Q06spark-tpch: 1 - Q05spark-tpch: 1 - Q04spark-tpch: 1 - Q03spark-tpch: 1 - Q02spark-tpch: 1 - Q01lczero: Eigenlczero: BLASab8.966.40176.670178.9106.5121.2910.1623.5721.162.449649162.329.48165.1044.7637209.79980.83123199.08.902.322809.01123041.610.7215020813358.5321.4244.8022208.1200635.8108.24886.4342608.00908.9336.750821.0031859.7248.5770224.57984.45035.5955131141.9715.036217108.4634122.0312571.87548.491720.6157123352.815.318365.20703984.62784.5178132.6580132.0485420.7448.447620.634591630925473719.2814796.0713383.2004190.79995.2377120.206554.409732.022931.21541761.4041281869895760123558.619.587458071758.593198622.054.50641.2413804.17841631.423244390.34.71264.7188212.0955211.74481703.42156.41591136.710584.251968.265514.642217.30235540.6268607.56642.320.110.4510.6932563887.8952891020.7987613710.4528725934.5130564324.3092791214.215703979.7773386612.7045501112.7590141319.4053777113.5802825324.3600374836.6645851126.7353528324.857111615.9030920729.8362789120.9959831226.1878871914.2548720012.007956196.0541189532.9071502711.435608236.2067783718.4697119412.770445506.871312945.841380767.076223697.377283739.944004388.0029234915.1748809821.9067020415.5182476114.652009332.0510474516.4436562912.3457120313.973087637.431042837.588891511.007690479.645312313.057396170.790923955.628538452.959939241.381472592.501859662.064853311.588159362.175426481.273381353.813596655.709694072.655846444.010448060.468229154.131221613.925257453.864421842.061790714.320060817048539.656.74184.347186.6096.2820.6010.4323.1221.572.495177472.289.63162.5614.829206.9690.82122070.38.982.342792.09123777.710.6579394213434.0921.3134.7775209.1955639.0888.20886.8412596.09618.9736.914621.0931728.9249.4983225.40474.43415.6159131613.6717.593617047.639121.6067569.95548.33220.683712297115.365365.00793996.76786.8905132.2719132.4219421.9148.326420.68691835961470717.9791797.4124382.5561191.09715.2296120.037354.485931.979531.25751759.0746282211175400123411.119.564756581756.556998528.854.55461.2404804.75281632.453243345.24.71174.718212.1305211.77291703.25156.42831136.643984.249168.263614.642617.30195540.517607.57352.320.110.4510.8741006977.7067565921.0538444512.0859279633.7419815124.5578899414.975358019.4828777312.5676708213.0449647917.7000179313.3120002724.6858558736.6652679426.6290950825.860551835.8838248331.2005977621.816757229.6859054614.5304679912.868350036.0443091432.7015495311.539661416.060416717.3137016313.013741496.952706815.438705926.906023037.9408378610.038290028.2781429314.7771949822.5255279514.5576934814.896059041.855959318.8612918911.2624216114.289846427.392459877.288267141.066792139.559095383.050016880.857975965.131711482.883481981.517799142.587141752.211469651.740741612.266416071.139986873.812455425.897758482.609078173.877902750.358015573.692176343.754272463.866108182.082242014.44657946715871OpenBenchmarking.org

PyTorch

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: ResNet-152ab3691215SE +/- 0.05, N = 38.969.65MIN: 4.84 / MAX: 9.24MIN: 4.98 / MAX: 9.85

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: Efficientnet_v2_lab246810SE +/- 0.05, N = 36.406.74MIN: 2.93 / MAX: 6.73MIN: 3.48 / MAX: 6.89

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 13 - Input: Bosphorus 4Kab4080120160200SE +/- 1.61, N = 15176.67184.351. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 12 - Input: Bosphorus 4Kab4080120160200SE +/- 1.43, N = 3178.91186.611. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

WebP2 Image Encode

This is a test of Google's libwebp2 library with the WebP2 image encode utility and using a sample 6000x4000 pixel JPEG image as the input, similar to the WebP/libwebp test profile. WebP2 is currently experimental and under heavy development as ultimately the successor to WebP. WebP2 supports 10-bit HDR, more efficienct lossy compression, improved lossless compression, animation support, and full multi-threading support compared to WebP. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP2 Image Encode 20220823Encode Settings: Quality 100, Compression Effort 5ab246810SE +/- 0.04, N = 36.516.281. (CXX) g++ options: -msse4.2 -fno-rtti -O3 -ldl

PyTorch

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: ResNet-50ab510152025SE +/- 0.31, N = 321.2920.60MIN: 13.22 / MAX: 22.39MIN: 13.89 / MAX: 21.35

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: ResNet-152ab3691215SE +/- 0.08, N = 310.1610.43MIN: 4.56 / MAX: 10.94MIN: 4.8 / MAX: 11.36

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: ResNet-50ab612182430SE +/- 0.19, N = 1523.5723.12MIN: 11.38 / MAX: 25.62MIN: 12.17 / MAX: 24.33

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: ResNet-50ab510152025SE +/- 0.25, N = 321.1621.57MIN: 12.26 / MAX: 22.24MIN: 14.06 / MAX: 22.29

Apache Spark TPC-H

This is a benchmark of Apache Spark using TPC-H data-set. Apache Spark is an open-source unified analytics engine for large-scale data processing and dealing with big data. This test profile benchmarks the Apache Spark in a single-system configuration using spark-submit. The test makes use of https://github.com/ssavvides/tpch-spark/ for facilitating the TPC-H benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 1 - Geometric Mean Of All Queriesab0.56141.12281.68422.24562.807SE +/- 0.02040294, N = 32.449649162.49517747MIN: 0.73 / MAX: 10.03MIN: 0.86 / MAX: 9.56

PyTorch

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 256 - Model: Efficientnet_v2_lab0.5221.0441.5662.0882.61SE +/- 0.00, N = 32.322.28MIN: 1.83 / MAX: 2.8MIN: 1.71 / MAX: 2.84

WebP2 Image Encode

This is a test of Google's libwebp2 library with the WebP2 image encode utility and using a sample 6000x4000 pixel JPEG image as the input, similar to the WebP/libwebp test profile. WebP2 is currently experimental and under heavy development as ultimately the successor to WebP. WebP2 supports 10-bit HDR, more efficienct lossy compression, improved lossless compression, animation support, and full multi-threading support compared to WebP. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP2 Image Encode 20220823Encode Settings: Defaultab3691215SE +/- 0.08, N = 39.489.631. (CXX) g++ options: -msse4.2 -fno-rtti -O3 -ldl

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 8 - Input: Bosphorus 1080pab4080120160200SE +/- 1.87, N = 3165.10162.561. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: ResNet-50, Baseline - Scenario: Synchronous Single-Streamab1.08652.1733.25954.3465.4325SE +/- 0.0118, N = 34.76374.8290

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: ResNet-50, Baseline - Scenario: Synchronous Single-Streamab50100150200250SE +/- 0.52, N = 3209.80206.97

WebP2 Image Encode

This is a test of Google's libwebp2 library with the WebP2 image encode utility and using a sample 6000x4000 pixel JPEG image as the input, similar to the WebP/libwebp test profile. WebP2 is currently experimental and under heavy development as ultimately the successor to WebP. WebP2 supports 10-bit HDR, more efficienct lossy compression, improved lossless compression, animation support, and full multi-threading support compared to WebP. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP2 Image Encode 20220823Encode Settings: Quality 75, Compression Effort 7ab0.18680.37360.56040.74720.934SE +/- 0.00, N = 30.830.821. (CXX) g++ options: -msse4.2 -fno-rtti -O3 -ldl

Xmrig

Xmrig is an open-source cross-platform CPU/GPU miner for RandomX, KawPow, CryptoNight and AstroBWT. This test profile is setup to measure the Xmrig CPU mining performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgH/s, More Is BetterXmrig 6.21Variant: CryptoNight-Femto UPX2 - Hash Count: 1Mab30K60K90K120K150KSE +/- 220.87, N = 3123199.0122070.31. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc

PyTorch

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 32 - Model: ResNet-152ab3691215SE +/- 0.10, N = 38.908.98MIN: 4.8 / MAX: 9.23MIN: 5.1 / MAX: 9.29

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: Efficientnet_v2_lab0.52651.0531.57952.1062.6325SE +/- 0.01, N = 32.322.34MIN: 1.77 / MAX: 2.81MIN: 1.78 / MAX: 2.78

Java SciMark

This test runs the Java version of SciMark 2, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.2Computational Test: Sparse Matrix Multiplyab6001200180024003000SE +/- 3.16, N = 32809.012792.09

Xmrig

Xmrig is an open-source cross-platform CPU/GPU miner for RandomX, KawPow, CryptoNight and AstroBWT. This test profile is setup to measure the Xmrig CPU mining performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgH/s, More Is BetterXmrig 6.21Variant: CryptoNight-Heavy - Hash Count: 1Mab30K60K90K120K150KSE +/- 33.09, N = 3123041.6123777.71. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc

Apache Spark TPC-H

This is a benchmark of Apache Spark using TPC-H data-set. Apache Spark is an open-source unified analytics engine for large-scale data processing and dealing with big data. This test profile benchmarks the Apache Spark in a single-system configuration using spark-submit. The test makes use of https://github.com/ssavvides/tpch-spark/ for facilitating the TPC-H benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 10 - Geometric Mean Of All Queriesab3691215SE +/- 0.02, N = 310.7210.66MIN: 5.7 / MAX: 33.03MIN: 5.44 / MAX: 32.7

Java SciMark

This test runs the Java version of SciMark 2, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.2Computational Test: Dense LU Matrix Factorizationab3K6K9K12K15KSE +/- 31.70, N = 313358.5313434.09

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 4 - Input: Bosphorus 1080pab510152025SE +/- 0.13, N = 321.4221.311. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Streamab1.08052.1613.24154.3225.4025SE +/- 0.0103, N = 34.80224.7775

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Streamab50100150200250SE +/- 0.44, N = 3208.12209.20

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 13 - Input: Bosphorus 1080pab140280420560700SE +/- 8.75, N = 3635.81639.091. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 4 - Input: Bosphorus 4Kab246810SE +/- 0.041, N = 38.2488.2081. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 8 - Input: Bosphorus 4Kab20406080100SE +/- 0.17, N = 386.4386.841. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-Streamab6001200180024003000SE +/- 6.37, N = 32608.012596.10

PyTorch

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: ResNet-152ab3691215SE +/- 0.10, N = 38.938.97MIN: 4.75 / MAX: 9.39MIN: 4.96 / MAX: 9.11

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-Streamab816243240SE +/- 0.09, N = 336.7536.91

PyTorch

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 32 - Model: ResNet-50ab510152025SE +/- 0.20, N = 321.0021.09MIN: 11.39 / MAX: 21.87MIN: 13.93 / MAX: 21.71

Xmrig

Xmrig is an open-source cross-platform CPU/GPU miner for RandomX, KawPow, CryptoNight and AstroBWT. This test profile is setup to measure the Xmrig CPU mining performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgH/s, More Is BetterXmrig 6.21Variant: GhostRider - Hash Count: 1Mab7K14K21K28K35KSE +/- 24.02, N = 331859.731728.91. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Streamab50100150200250SE +/- 0.41, N = 3248.58249.50

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Streamab50100150200250SE +/- 0.01, N = 3224.58225.40

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Streamab1.00132.00263.00394.00525.0065SE +/- 0.0001, N = 34.45034.4341

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-Streamab1.26362.52723.79085.05446.318SE +/- 0.0055, N = 35.59555.6159

Xmrig

Xmrig is an open-source cross-platform CPU/GPU miner for RandomX, KawPow, CryptoNight and AstroBWT. This test profile is setup to measure the Xmrig CPU mining performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgH/s, More Is BetterXmrig 6.21Variant: Wownero - Hash Count: 1Mab30K60K90K120K150KSE +/- 621.69, N = 3131141.9131613.61. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Streamab150300450600750SE +/- 4.21, N = 3715.04717.59

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-Streamab4K8K12K16K20KSE +/- 16.76, N = 317108.4617047.64

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Streamab306090120150SE +/- 0.25, N = 3122.03121.61

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.8Encoder Mode: Preset 12 - Input: Bosphorus 1080pab120240360480600SE +/- 1.39, N = 3571.88569.961. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Streamab1122334455SE +/- 0.05, N = 348.4948.33

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Streamab510152025SE +/- 0.02, N = 320.6220.68

Xmrig

Xmrig is an open-source cross-platform CPU/GPU miner for RandomX, KawPow, CryptoNight and AstroBWT. This test profile is setup to measure the Xmrig CPU mining performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgH/s, More Is BetterXmrig 6.21Variant: Monero - Hash Count: 1Mab30K60K90K120K150KSE +/- 404.54, N = 3123352.8122971.01. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Streamab48121620SE +/- 0.01, N = 315.3215.37

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Streamab1530456075SE +/- 0.04, N = 365.2165.01

Java SciMark

This test runs the Java version of SciMark 2, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.2Computational Test: Compositeab9001800270036004500SE +/- 6.24, N = 33984.623996.76

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Streamab2004006008001000SE +/- 1.42, N = 3784.52786.89

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Streamab306090120150SE +/- 0.66, N = 3132.66132.27

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Streamab306090120150SE +/- 0.03, N = 3132.05132.42

Java SciMark

This test runs the Java version of SciMark 2, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.2Computational Test: Fast Fourier Transformab90180270360450SE +/- 0.36, N = 3420.74421.91

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Streamab1122334455SE +/- 0.02, N = 348.4548.33

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Streamab510152025SE +/- 0.01, N = 320.6320.69

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. The system/openssl test profiles relies on benchmarking the system/OS-supplied openssl binary rather than the pts/openssl test profile that uses the locally-built OpenSSL for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSLAlgorithm: SHA512ab20000M40000M60000M80000M100000MSE +/- 191332047.54, N = 391630925473918359614701. OpenSSL 3.0.10 1 Aug 2023 (Library: OpenSSL 3.0.10 1 Aug 2023)

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Streamab160320480640800SE +/- 1.53, N = 3719.28717.98

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-Streamab2004006008001000SE +/- 1.54, N = 3796.07797.41

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Streamab80160240320400SE +/- 0.61, N = 3383.20382.56

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Synchronous Single-Streamab4080120160200SE +/- 0.06, N = 3190.80191.10

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Synchronous Single-Streamab1.17852.3573.53554.7145.8925SE +/- 0.0015, N = 35.23775.2296

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-Streamab306090120150SE +/- 0.24, N = 3120.21120.04

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Streamab1224364860SE +/- 0.07, N = 354.4154.49

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: BERT-Large, NLP Question Answering - Scenario: Synchronous Single-Streamab714212835SE +/- 0.03, N = 332.0231.98

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: BERT-Large, NLP Question Answering - Scenario: Synchronous Single-Streamab714212835SE +/- 0.03, N = 331.2231.26

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Streamab400800120016002000SE +/- 2.24, N = 31761.401759.07

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. The system/openssl test profiles relies on benchmarking the system/OS-supplied openssl binary rather than the pts/openssl test profile that uses the locally-built OpenSSL for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSLAlgorithm: SHA256ab60000M120000M180000M240000M300000MSE +/- 548972949.20, N = 32818698957602822111754001. OpenSSL 3.0.10 1 Aug 2023 (Library: OpenSSL 3.0.10 1 Aug 2023)

Xmrig

Xmrig is an open-source cross-platform CPU/GPU miner for RandomX, KawPow, CryptoNight and AstroBWT. This test profile is setup to measure the Xmrig CPU mining performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgH/s, More Is BetterXmrig 6.21Variant: KawPow - Hash Count: 1Mab30K60K90K120K150KSE +/- 87.00, N = 3123558.6123411.11. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc

Apache Spark TPC-H

This is a benchmark of Apache Spark using TPC-H data-set. Apache Spark is an open-source unified analytics engine for large-scale data processing and dealing with big data. This test profile benchmarks the Apache Spark in a single-system configuration using spark-submit. The test makes use of https://github.com/ssavvides/tpch-spark/ for facilitating the TPC-H benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterApache Spark TPC-H 3.5Scale Factor: 50 - Geometric Mean Of All Queriesab510152025SE +/- 0.05, N = 319.5919.56MIN: 9.71 / MAX: 103.64MIN: 9.48 / MAX: 77.71

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-Streamab400800120016002000SE +/- 1.91, N = 31758.591756.56

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. The system/openssl test profiles relies on benchmarking the system/OS-supplied openssl binary rather than the pts/openssl test profile that uses the locally-built OpenSSL for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsign/s, More Is BetterOpenSSLAlgorithm: RSA4096ab20K40K60K80K100KSE +/- 53.45, N = 398622.098528.81. OpenSSL 3.0.10 1 Aug 2023 (Library: OpenSSL 3.0.10 1 Aug 2023)

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-Streamab1224364860SE +/- 0.06, N = 354.5154.55

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: ResNet-50, Sparse INT8 - Scenario: Synchronous Single-Streamab0.27930.55860.83791.11721.3965SE +/- 0.0046, N = 31.24131.2404

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: ResNet-50, Sparse INT8 - Scenario: Synchronous Single-Streamab2004006008001000SE +/- 3.00, N = 3804.18804.75

Java SciMark

This test runs the Java version of SciMark 2, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.2Computational Test: Monte Carloab400800120016002000SE +/- 0.75, N = 31631.421632.45

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. The system/openssl test profiles relies on benchmarking the system/OS-supplied openssl binary rather than the pts/openssl test profile that uses the locally-built OpenSSL for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgverify/s, More Is BetterOpenSSLAlgorithm: RSA4096ab700K1400K2100K2800K3500KSE +/- 1292.47, N = 33244390.33243345.21. OpenSSL 3.0.10 1 Aug 2023 (Library: OpenSSL 3.0.10 1 Aug 2023)

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Synchronous Single-Streamab1.06032.12063.18094.24125.3015SE +/- 0.0079, N = 34.71264.7117

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: CV Detection, YOLOv5s COCO - Scenario: Synchronous Single-Streamab1.06172.12343.18514.24685.3085SE +/- 0.0110, N = 34.71884.7180

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Synchronous Single-Streamab50100150200250SE +/- 0.35, N = 3212.10212.13

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: CV Detection, YOLOv5s COCO - Scenario: Synchronous Single-Streamab50100150200250SE +/- 0.50, N = 3211.74211.77

Java SciMark

This test runs the Java version of SciMark 2, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.2Computational Test: Jacobi Successive Over-Relaxationab400800120016002000SE +/- 0.16, N = 31703.421703.25

Neural Magic DeepSparse

This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: BERT-Large, NLP Question Answering - Scenario: Asynchronous Multi-Streamab306090120150SE +/- 0.02, N = 3156.42156.43

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Streamab2004006008001000SE +/- 2.45, N = 31136.711136.64

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Streamab20406080100SE +/- 0.19, N = 384.2584.25

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Synchronous Single-Streamab1530456075SE +/- 0.11, N = 368.2768.26

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Synchronous Single-Streamab48121620SE +/- 0.02, N = 314.6414.64

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-Streamab48121620SE +/- 0.01, N = 317.3017.30

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.6Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-Streamab12002400360048006000SE +/- 5.02, N = 35540.635540.52

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.6Model: BERT-Large, NLP Question Answering - Scenario: Asynchronous Multi-Streamab130260390520650SE +/- 0.37, N = 3607.57607.57

PyTorch

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 32 - Model: Efficientnet_v2_lab0.5221.0441.5662.0882.61SE +/- 0.01, N = 32.322.32MIN: 1.86 / MAX: 2.8MIN: 1.93 / MAX: 2.71

WebP2 Image Encode

This is a test of Google's libwebp2 library with the WebP2 image encode utility and using a sample 6000x4000 pixel JPEG image as the input, similar to the WebP/libwebp test profile. WebP2 is currently experimental and under heavy development as ultimately the successor to WebP. WebP2 supports 10-bit HDR, more efficienct lossy compression, improved lossless compression, animation support, and full multi-threading support compared to WebP. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP2 Image Encode 20220823Encode Settings: Quality 100, Lossless Compressionab0.02480.04960.07440.09920.124SE +/- 0.00, N = 30.110.111. (CXX) g++ options: -msse4.2 -fno-rtti -O3 -ldl

OpenBenchmarking.orgMP/s, More Is BetterWebP2 Image Encode 20220823Encode Settings: Quality 95, Compression Effort 7ab0.10130.20260.30390.40520.5065SE +/- 0.00, N = 30.450.451. (CXX) g++ options: -msse4.2 -fno-rtti -O3 -ldl

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. The system/openssl test profiles relies on benchmarking the system/OS-supplied openssl binary rather than the pts/openssl test profile that uses the locally-built OpenSSL for benchmarking. Learn more via the OpenBenchmarking.org test page.

Algorithm: ChaCha20-Poly1305

a: The test run did not produce a result.

b: The test run did not produce a result.

Algorithm: AES-256-GCM

a: The test run did not produce a result. E: 40270E64087F0000:error:1C800066:Provider routines:ossl_gcm_stream_update:cipher operation failed:../providers/implementations/ciphers/ciphercommon_gcm.c:320:

b: The test run did not produce a result. E: 408712FE017F0000:error:1C800066:Provider routines:ossl_gcm_stream_update:cipher operation failed:../providers/implementations/ciphers/ciphercommon_gcm.c:320:

Algorithm: AES-128-GCM

a: The test run did not produce a result. E: 4097A6F7B77F0000:error:1C800066:Provider routines:ossl_gcm_stream_update:cipher operation failed:../providers/implementations/ciphers/ciphercommon_gcm.c:320:

b: The test run did not produce a result. E: 40B7EFA3BE7F0000:error:1C800066:Provider routines:ossl_gcm_stream_update:cipher operation failed:../providers/implementations/ciphers/ciphercommon_gcm.c:320:

Algorithm: ChaCha20

a: The test run did not produce a result.

b: The test run did not produce a result.

LeelaChessZero

LeelaChessZero (lc0 / lczero) is a chess engine automated vian neural networks. This test profile can be used for OpenCL, CUDA + cuDNN, and BLAS (CPU-based) benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.30Backend: Eigenab150300450600750SE +/- 17.59, N = 87047151. (CXX) g++ options: -flto -pthread

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.30Backend: BLASab2004006008001000SE +/- 18.54, N = 98538711. (CXX) g++ options: -flto -pthread

94 Results Shown

PyTorch:
  CPU - 256 - ResNet-152
  CPU - 1 - Efficientnet_v2_l
SVT-AV1:
  Preset 13 - Bosphorus 4K
  Preset 12 - Bosphorus 4K
WebP2 Image Encode
PyTorch:
  CPU - 256 - ResNet-50
  CPU - 1 - ResNet-152
  CPU - 1 - ResNet-50
  CPU - 16 - ResNet-50
Apache Spark TPC-H
PyTorch
WebP2 Image Encode
SVT-AV1
Neural Magic DeepSparse:
  ResNet-50, Baseline - Synchronous Single-Stream:
    ms/batch
    items/sec
WebP2 Image Encode
Xmrig
PyTorch:
  CPU - 32 - ResNet-152
  CPU - 16 - Efficientnet_v2_l
Java SciMark
Xmrig
Apache Spark TPC-H
Java SciMark
SVT-AV1
Neural Magic DeepSparse:
  CV Classification, ResNet-50 ImageNet - Synchronous Single-Stream:
    ms/batch
    items/sec
SVT-AV1:
  Preset 13 - Bosphorus 1080p
  Preset 4 - Bosphorus 4K
  Preset 8 - Bosphorus 4K
Neural Magic DeepSparse
PyTorch
Neural Magic DeepSparse
PyTorch
Xmrig
Neural Magic DeepSparse:
  CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Stream
  NLP Text Classification, DistilBERT mnli - Synchronous Single-Stream
  NLP Text Classification, DistilBERT mnli - Synchronous Single-Stream
  ResNet-50, Sparse INT8 - Asynchronous Multi-Stream
Xmrig
Neural Magic DeepSparse:
  NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream
  ResNet-50, Sparse INT8 - Asynchronous Multi-Stream
  CV Detection, YOLOv5s COCO - Asynchronous Multi-Stream
SVT-AV1
Neural Magic DeepSparse:
  NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Stream:
    items/sec
    ms/batch
Xmrig
Neural Magic DeepSparse:
  CV Segmentation, 90% Pruned YOLACT Pruned - Synchronous Single-Stream:
    ms/batch
    items/sec
Java SciMark
Neural Magic DeepSparse:
  CV Detection, YOLOv5s COCO - Asynchronous Multi-Stream
  NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream
  NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream
Java SciMark
Neural Magic DeepSparse:
  NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Stream:
    items/sec
    ms/batch
OpenSSL
Neural Magic DeepSparse:
  NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream
  CV Detection, YOLOv5s COCO, Sparse INT8 - Asynchronous Multi-Stream
  CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Stream
  NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Synchronous Single-Stream
  NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Synchronous Single-Stream
  CV Detection, YOLOv5s COCO, Sparse INT8 - Asynchronous Multi-Stream
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream
  BERT-Large, NLP Question Answering - Synchronous Single-Stream
  BERT-Large, NLP Question Answering - Synchronous Single-Stream
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream
OpenSSL
Xmrig
Apache Spark TPC-H
Neural Magic DeepSparse
OpenSSL
Neural Magic DeepSparse:
  ResNet-50, Baseline - Asynchronous Multi-Stream
  ResNet-50, Sparse INT8 - Synchronous Single-Stream
  ResNet-50, Sparse INT8 - Synchronous Single-Stream
Java SciMark
OpenSSL
Neural Magic DeepSparse:
  CV Detection, YOLOv5s COCO, Sparse INT8 - Synchronous Single-Stream
  CV Detection, YOLOv5s COCO - Synchronous Single-Stream
  CV Detection, YOLOv5s COCO, Sparse INT8 - Synchronous Single-Stream
  CV Detection, YOLOv5s COCO - Synchronous Single-Stream
Java SciMark
Neural Magic DeepSparse:
  BERT-Large, NLP Question Answering - Asynchronous Multi-Stream
  NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream
  NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream
  BERT-Large, NLP Question Answering, Sparse INT8 - Synchronous Single-Stream
  BERT-Large, NLP Question Answering, Sparse INT8 - Synchronous Single-Stream
  NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Stream
  NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Stream
  BERT-Large, NLP Question Answering - Asynchronous Multi-Stream
PyTorch
WebP2 Image Encode:
  Quality 100, Lossless Compression
  Quality 95, Compression Effort 7
LeelaChessZero:
  Eigen
  BLAS