AMD AOCC 4.0 Benchmarks

AMD Ryzen 9 7950X compiler benchmarks by Michael Larabel for a future article.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2211152-PTS-AMDAOCC460
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:

Audio Encoding 2 Tests
AV1 3 Tests
BLAS (Basic Linear Algebra Sub-Routine) Tests 2 Tests
C++ Boost Tests 3 Tests
Chess Test Suite 2 Tests
C/C++ Compiler Tests 19 Tests
Compression Tests 2 Tests
CPU Massive 22 Tests
Creator Workloads 26 Tests
Cryptography 4 Tests
Database Test Suite 3 Tests
Encoding 8 Tests
Fortran Tests 3 Tests
Game Development 3 Tests
HPC - High Performance Computing 11 Tests
Imaging 7 Tests
Common Kernel Benchmarks 2 Tests
Machine Learning 7 Tests
Molecular Dynamics 2 Tests
MPI Benchmarks 2 Tests
Multi-Core 19 Tests
NVIDIA GPU Compute 3 Tests
Intel oneAPI 2 Tests
OpenMPI Tests 2 Tests
Programmer / Developer System Benchmarks 4 Tests
Python Tests 3 Tests
Raytracing 4 Tests
Renderers 5 Tests
Scientific Computing 4 Tests
Server 5 Tests
Server CPU Tests 10 Tests
Single-Threaded 6 Tests
Telephony 2 Tests
Texture Compression 3 Tests
Video Encoding 6 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
AOCC 4.0
November 12 2022
  8 Hours, 37 Minutes
GCC 12.2
November 13 2022
  9 Hours, 30 Minutes
GCC 13.0 14 Nov
November 14 2022
  5 Hours, 9 Minutes
LLVM Clang 14
November 13 2022
  9 Hours, 10 Minutes
LLVM Clang 15.0.2
November 14 2022
  7 Hours, 48 Minutes
Invert Hiding All Results Option
  8 Hours, 3 Minutes

Only show results where is faster than
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):


AMD AOCC 4.0 BenchmarksOpenBenchmarking.orgPhoronix Test SuiteAMD Ryzen 9 7950X 16-Core @ 5.88GHz (16 Cores / 32 Threads)ASUS ROG CROSSHAIR X670E HERO (0703 BIOS)AMD Device 14d832GB1000GB Sabrent Rocket 4.0 PlusAMD Radeon RX 6800 16GB (2475/1000MHz)AMD Navi 21/23ASUS MG28UIntel I225-V + Intel Wi-Fi 6 AX210/AX211/AX411Ubuntu 22.106.1.0-060100rc3daily20221103-generic (x86_64)GNOME Shell 43.0X Server + Wayland4.6 Mesa 22.2.1 (LLVM 15.0.2 DRM 3.49)1.3.224Clang 14.0.6GCC 12.2.0GCC 13.0.0 20221114 + clang (GCC) 13.0.0 20221114 (experimental)Clang 14.0.6-2Clang 15.0.2-1ext43840x2160ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerOpenGLVulkanCompilersFile-SystemScreen ResolutionAMD AOCC 4.0 Benchmarks PerformanceSystem Logs- Transparent Huge Pages: madvise- CXXFLAGS="-O3 -march=native -flto" CFLAGS="-O3 -march=native -flto"- AOCC 4.0: Optimized build with assertions; Default target: x86_64-unknown-linux-gnu; Host CPU: znver4 - GCC 12.2: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --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-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-U8K4Qv/gcc-12-12.2.0/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-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 - GCC 13.0 14 Nov: --disable-multilib- Scaling Governor: amd-pstate schedutil (Boost: Enabled) - CPU Microcode: 0xa601203 - BAR1 / Visible vRAM Size: 16368 MB - vBIOS Version: 113-D4120900-101- Python 3.10.7- itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + 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 IBRS_FW STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

AOCC 4.0GCC 12.2GCC 13.0 14 NovLLVM Clang 14LLVM Clang 15.0.2Result OverviewPhoronix Test Suite100%150%200%250%300%CaffeOpenJPEGC-RayLiquid-DSPKripkeJPEG XL libjxlTSCPCoremarkJPEG XL Decoding libjxlASTC EncoderFLAC Audio EncodingNCNNXsbenchlibavif avifencAOM AV1CLOMPSecureMarkAOBenchWebP Image EncodeKvazaarNgspiceLAME MP3 EncodingC-BlosconeDNNGraphicsMagickSQLite SpeedtestLAMMPS Molecular Dynamics SimulatorQuadRayCppPerformanceBenchmarksOpenSSLMonte Carlo Simulations of Ionised NebulaeDolfynSVT-AV1libjpeg-turbo tjbenchSockperfKTX-Software toktxSVT-HEVCNettleZstd CompressionSVT-VP9Tachyon

AMD AOCC 4.0 Benchmarksjpegxl: JPEG - 100onnx: ArcFace ResNet-100 - CPU - Standardlczero: BLASlczero: Eigencryptopp: Keyed Algorithmssecuremark: SecureMark-TLSjpegxl: PNG - 100tnn: CPU - DenseNetonnx: GPT-2 - CPU - Standardonnx: yolov4 - CPU - Standardopenssl: SHA256cpp-perf-bench: Math Libraryonnx: super-resolution-10 - CPU - Parallelonnx: GPT-2 - CPU - Parallelonnx: bertsquad-12 - CPU - Parallelonnx: yolov4 - CPU - Parallelonnx: ArcFace ResNet-100 - CPU - Parallelsimdjson: Kostyaonnx: fcn-resnet101-11 - CPU - Parallelsvt-hevc: 1 - Bosphorus 4Kncnn: CPU - regnety_400mncnn: CPU - squeezenet_ssdncnn: CPU - yolov4-tinyncnn: CPU - resnet50ncnn: CPU - alexnetncnn: CPU - resnet18ncnn: CPU - vgg16ncnn: CPU - googlenetncnn: CPU - blazefacencnn: CPU - efficientnet-b0ncnn: CPU - mnasnetncnn: CPU - shufflenet-v2ncnn: CPU-v3-v3 - mobilenet-v3ncnn: CPU-v2-v2 - mobilenet-v2ncnn: CPU - mobilenetcaffe: GoogleNet - CPU - 100simdjson: LargeRandonednn: Recurrent Neural Network Training - u8s8f32 - CPUonednn: Recurrent Neural Network Training - bf16bf16bf16 - CPUonednn: Recurrent Neural Network Inference - bf16bf16bf16 - CPUonednn: Recurrent Neural Network Inference - u8s8f32 - CPUjpegxl: JPEG - 90jpegxl: PNG - 90mocassin: Dust 2D tau100.0avifenc: 0openvino: Person Vehicle Bike Detection FP16 - CPUopenvino: Person Vehicle Bike Detection FP16 - CPUsimdjson: DistinctUserIDsimdjson: PartialTweetssimdjson: TopTweetopenvino: Person Detection FP16 - CPUopenvino: Person Detection FP16 - CPUopenvino: Person Detection FP32 - CPUopenvino: Person Detection FP32 - CPUopenvino: Face Detection FP16 - CPUopenvino: Face Detection FP16 - CPUopenvino: Face Detection FP16-INT8 - CPUopenvino: Face Detection FP16-INT8 - CPUopenvino: Machine Translation EN To DE FP16 - CPUopenvino: Machine Translation EN To DE FP16 - CPUngspice: C2670openvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUopenvino: Age Gender Recognition Retail 0013 FP16 - CPUopenvino: Age Gender Recognition Retail 0013 FP16 - CPUopenvino: Vehicle Detection FP16-INT8 - CPUopenvino: Vehicle Detection FP16-INT8 - CPUopenvino: Vehicle Detection FP16 - CPUopenvino: Vehicle Detection FP16 - CPUopenvino: Weld Porosity Detection FP16 - CPUopenvino: Weld Porosity Detection FP16 - CPUopenvino: Weld Porosity Detection FP16-INT8 - CPUopenvino: Weld Porosity Detection FP16-INT8 - CPUgraphics-magick: Sharpengraphics-magick: Noise-Gaussiangraphics-magick: Resizinggraphics-magick: Enhancedgraphics-magick: Rotateopenssl: RSA4096openssl: RSA4096graphics-magick: Swirlgraphics-magick: HWB Color Spacengspice: C7552tachyon: Total Timesvt-av1: Preset 4 - Bosphorus 4Kredis: SET - 50cryptopp: Unkeyed Algorithmskripke: pjsip: OPTIONS, Statefulpjsip: INVITEcompress-zstd: 19 - Decompression Speedcompress-zstd: 19 - Compression Speeddragonflydb: 50 - 1:5dragonflydb: 50 - 5:1aom-av1: Speed 6 Two-Pass - Bosphorus 4Kjpegxl-decode: 1quantlib: cpp-perf-bench: Stepanov Vectorcompress-zstd: 19, Long Mode - Decompression Speedcompress-zstd: 19, Long Mode - Compression Speedonednn: IP Shapes 1D - bf16bf16bf16 - CPUsqlite-speedtest: Timed Time - Size 1,000caffe: AlexNet - CPU - 100compress-zstd: 3 - Decompression Speedcompress-zstd: 3 - Compression Speedavifenc: 2cpp-perf-bench: Ctypeliquid-dsp: 8 - 256 - 57svt-vp9: VMAF Optimized - Bosphorus 4Kwebp: Quality 100, Lossless, Highest Compressioncpp-perf-bench: Atolonednn: IP Shapes 1D - u8s8f32 - CPUcoremark: CoreMark Size 666 - Iterations Per Secondastcenc: Exhaustivetjbench: Decompression Throughputtnn: CPU - MobileNet v2onednn: IP Shapes 3D - u8s8f32 - CPUespeak: Text-To-Speech Synthesisjpegxl-decode: Allc-ray: Total Time - 4K, 16 Rays Per Pixelonednn: Deconvolution Batch shapes_1d - bf16bf16bf16 - CPUonednn: Deconvolution Batch shapes_1d - u8s8f32 - CPUaobench: 2048 x 2048 - Total Timeliquid-dsp: 32 - 256 - 57liquid-dsp: 16 - 256 - 57tnn: CPU - SqueezeNet v1.1quadray: 5 - 4Kquadray: 1 - 4Kquadray: 5 - 1080pquadray: 1 - 1080pclomp: Static OMP Speeduppovray: Trace Timeencode-flac: WAV To FLACcpp-perf-bench: Stepanov Abstractionastcenc: Thoroughtnn: CPU - SqueezeNet v2sockperf: Throughputpjsip: OPTIONS, Statelesskvazaar: Bosphorus 4K - Very Fastblosc: blosclz bitshuffleonednn: Matrix Multiply Batch Shapes Transformer - bf16bf16bf16 - CPUonednn: Matrix Multiply Batch Shapes Transformer - u8s8f32 - CPUastcenc: Fasttoktx: Zstd Compression 19webp: Quality 100, Losslessdolfyn: Computational Fluid Dynamicsopenjpeg: NASA Curiosity Panorama M34aom-av1: Speed 10 Realtime - Bosphorus 4Kxsbench: cpp-perf-bench: Function Objectsonednn: IP Shapes 3D - bf16bf16bf16 - CPUsvt-av1: Preset 8 - Bosphorus 4Kblosc: blosclz shufflekvazaar: Bosphorus 4K - Ultra Fastdraco: Church Facadesvt-vp9: Visual Quality Optimized - Bosphorus 4Ksvt-hevc: 7 - Bosphorus 4Kprimesieve: 1e12onednn: Convolution Batch Shapes Auto - u8s8f32 - CPUsvt-vp9: PSNR/SSIM Optimized - Bosphorus 4Kastcenc: Mediumkvazaar: Bosphorus 1080p - Very Fastsvt-av1: Preset 10 - Bosphorus 4Konednn: Deconvolution Batch shapes_3d - bf16bf16bf16 - CPUtoktx: UASTC 3webp: Quality 100, Highest Compressiondraco: Lionencode-mp3: WAV To MP3avifenc: 6, Losslessyquake2: Software CPU Color Light - Off - Off - 1920 x 1080lammps: Rhodopsin Proteinsvt-hevc: 10 - Bosphorus 4Knettle: aes256yquake2: Software CPU Color Light - On - Off - 1920 x 1080yquake2: Software CPU Color Light - On - On - 1920 x 1080svt-av1: Preset 12 - Bosphorus 4Kyquake2: Software CPU Color Light - Off - On - 1920 x 1080yquake2: Software CPU - On - On - 1920 x 1080yquake2: Software CPU - On - Off - 1920 x 1080yquake2: Software CPU - Off - Off - 1920 x 1080avifenc: 6avifenc: 10, Losslesskvazaar: Bosphorus 1080p - Ultra Fastyquake2: Software CPU - Off - On - 1920 x 1080nettle: sha512webp: Quality 100nettle: chachatscp: AI Chess Performancewebp: Defaultnettle: poly1305-aesAOCC 4.0GCC 12.2GCC 13.0 14 NovLLVM Clang 14LLVM Clang 15.0.20.852207174819451048.6478434759411.082455.642914757640426163787164.7999113713591756923565.851356.687.5511.2714.3811.624.797.1824.837.851.213.852.722.962.693.178.57739341.591138.831139.44574.905574.00813.1313.506463.1144.801664.7711.9711.8210.851055.297.531081.017.37557.0114.29283.0728.1959.95133.3358.2550.2563880.920.3644290.734.321847.7310.62752.275.551439.925.522896.904706072609612952394139.16026.21413171056.31255.48673.0784701871.5591.183634125045767937053445018.679.15728724.375242954.5920.7577.274532.339.3805117.854.60.57783434.774276535461.25092.231.37831.359896243333112.440.8925.5190.338581946032.2882052.0160318.287249244.0450.34701714.535286.4427.4533.273670.40509721.57918866000001754833333217.8352.0526.438.12102.7925.016.1029.92816.93219.906044.59287909512452347.3313508.90.2223780.133169427.333111.3462.1710.7677399165.4470813469.0231.5001676.51123837.285.354324111.78108.326.2455.28785121.78160.9599113.72131.2941.384554.9576.3333734.7324.229191.817.910171.098564.64188.2185.3188.400189.9196.3198.1205.72.9432.773219.33205.6864.2917.671528.82247706329.654255.910.85222216971725881.3523484328101.044513.705894056436864232800166.5759395712691856823615.891346.6412.0112.4314.0312.254.997.4924.568.661.594.643.303.763.243.718.68657421.831140.341139.55578.822580.33012.5412.986468.5234.821658.309.809.859.901060.877.501085.337.34557.3614.29284.3528.0959.47134.4263.7460.2563460.600.3644470.594.351837.1910.55757.515.591429.515.552879.683436642741639988393865.86014.61473153358.28255.48333.0674498344.35481.755234133327173933052794932.080.35707956.745275879.8319.3268.444501.135.1964974.454.60.72597834.225236855432.14978.634.84532.277834506667109.950.8825.4270.4799061066150.9452072.0071314.759071412.9440.35654617.832263.9018.9943.364120.42141919.76216806000001627233333388.4392.0826.488.22102.3122.815.95510.52015.72719.454969.92388910312582541.7813017.30.2285140.139862289.289011.6852.2310.7877122268.2462345089.2511.5437976.39823526.480.314424111.60106.576.1875.32780122.32132.5876102.77131.0431.386964.9584.4834864.4095.028176.417.264172.538521.41172.5171.2187.372176.2185.1185.9192.23.2023.077204.74192.0858.3016.611515.12249682128.024182.310.854328961.0337385343083169.4796.036.6311.8412.2714.1212.264.987.4924.398.651.584.573.323.743.263.728.68649871.801139.641142.22578.955582.00312.7613.016366.2669.289.6310.2262.6763416582625627912394339.86029.11454143860.29455.28093.1094675697.0132057600938953384988.982.25757620.865271140.4920.7266.3035.5085073.854.30.66290534.734236965485.84976.233.75133.773650876667110.490.8925.5450.5242831117817.2471502.0276322.4912510.35640717.769245.5419.3263.426860.41886019.902163316666712744333331.9326.307.65102.0622.610.81015.73219.772590031012402844.1213286.50.2279180.140349294.572411.6162.2810.5106960868.3063182629.2111.5330576.36024105.082.56111.61107.165.30501122.36135.1913108.11132.7461.405034.9614.604.3874.675185.317.044173.368490.46178.5181.3188.616184.4191.4192.3201.03.1052.947210.39199.0859.9616.801507.34249222828.184188.140.942781170218611086.8466414726241.073634.328898056639948818857164.3858949706292356323535.621356.649.4111.8014.6812.325.197.3625.968.541.374.253.023.252.993.449.091049421.561138.261142.95575.942575.60213.1013.196565.2284.831654.6610.7711.1510.601066.077.461092.877.29555.0014.37283.7228.1559.83133.5957.7110.2563331.040.3644085.254.321851.0910.62752.415.571434.105.542887.984665872705582896395258.16016.61382158056.21654.97293.020593.2098141097949004913.278.917.7662.294669.540.1795055.854.60.58337736.072469765463.55045.232.37831.627809823333111.350.8325.4620.343595981244.4862041.9814320.162344431.5620.33776020.543264.4820.1343.395810.40843221.70417236000001584066667302.0222.0426.238.07101.8923.716.5689.28017.03019.279661.10588375644.6213989.20.2241010.136343404.449611.6392.0610.7716441664.8267560019.2461.5117674.60824796.882.974782110.25105.556.1725.31930120.82155.0895107.85130.1071.407664.9826.1637634.6204.37717.688170.798566.17183.1913.0792.761210.45864.8717.511536.84294257929.454231.630.87168618631075.7906054549571.112499.82437916267870165.5466.639.6211.9714.9412.325.167.4125.528.631.414.353.073.312.993.539.241074141137.491142.00576.219575.12517.4617.986465.1504.851645.571065.317.471071.507.43555.0514.35282.5928.2759.20135.0058.9550.2563386.570.3644075.754.331845.3110.76742.445.561438.375.522895.024545832655583876395636.46030.91390158456.47755.08363.058596.8657831089772674894.678.817.9263.324783.940.2844928.955.30.58238136.017468005458.65030.832.31231.784566352750110.810.9026.0560.3437491004815.0069421.9886321.216455337.4810.362285266.8920.1533.464440.40859421.39316549333331544266667232.4351.9226.157.60102.3423.614.8349.43517.18319.430149.43389766145.3113694.20.2244490.136310412.209611.4572.1510.8226268968.7167522389.3331.5136974.81724197.784.804649111.24105.696.2435.30337121.26156.4532109.33129.1551.395794.9836.2137004.5294.36317.385171.028553.15184.1503.0442.725213.39863.7317.001523.44285702229.234214.74OpenBenchmarking.org

JPEG XL libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.7Input: JPEG - Quality: 100LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.21150.4230.63450.8461.0575SE +/- 0.01, N = 9SE +/- 0.01, N = 3SE +/- 0.02, N = 6SE +/- 0.01, N = 3SE +/- 0.01, N = 60.870.940.850.850.85-Xclang -mrelax-all-Xclang -mrelax-all-Xclang -mrelax-all1. (CXX) g++ options: -O3 -march=native -flto -fno-rtti -funwind-tables -O2 -fPIE -pie -latomic

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: ArcFace ResNet-100 - Device: CPU - Executor: StandardLLVM Clang 14GCC 12.2AOCC 4.06001200180024003000SE +/- 34.72, N = 12SE +/- 81.96, N = 12SE +/- 23.09, N = 12278122222207-flto=thin-flto=auto -fno-fat-lto-objects-flto=thin1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt

Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

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 0.28Backend: BLASLLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.0400800120016002000SE +/- 18.88, N = 4SE +/- 14.33, N = 3SE +/- 18.37, N = 4SE +/- 24.06, N = 316861702169717481. (CXX) g++ options: -flto -O3 -march=native -pthread

Backend: BLAS

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./lczero: line 4: ./lc0: No such file or directory

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.28Backend: EigenLLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.0400800120016002000SE +/- 11.93, N = 3SE +/- 17.46, N = 3SE +/- 19.08, N = 3SE +/- 19.60, N = 318631861172519451. (CXX) g++ options: -flto -O3 -march=native -pthread

Backend: Eigen

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./lczero: line 4: ./lc0: No such file or directory

Crypto++

Crypto++ is a C++ class library of cryptographic algorithms. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/second, More Is BetterCrypto++ 8.2Test: Keyed AlgorithmsLLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.02004006008001000SE +/- 3.44, N = 3SE +/- 0.29, N = 3SE +/- 1.11, N = 3SE +/- 2.11, N = 31075.791086.85881.351048.651. (CXX) g++ options: -O3 -march=native -flto -fPIC -pthread -pipe

Test: Keyed Algorithms

GCC 13.0 14 Nov: The test quit with a non-zero exit status.

SecureMark

SecureMark is an objective, standardized benchmarking framework for measuring the efficiency of cryptographic processing solutions developed by EEMBC. SecureMark-TLS is benchmarking Transport Layer Security performance with a focus on IoT/edge computing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmarks, More Is BetterSecureMark 1.0.4Benchmark: SecureMark-TLSLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0100K200K300K400K500KSE +/- 2486.77, N = 3SE +/- 535.99, N = 3SE +/- 1407.45, N = 3SE +/- 2377.52, N = 3SE +/- 454.53, N = 34549574726244328964328104759411. (CC) gcc options: -pedantic -O3

JPEG XL libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.7Input: PNG - Quality: 100LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.24980.49960.74940.99921.249SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 5SE +/- 0.01, N = 31.111.071.031.041.08-Xclang -mrelax-all-Xclang -mrelax-all-Xclang -mrelax-all1. (CXX) g++ options: -O3 -march=native -flto -fno-rtti -funwind-tables -O2 -fPIE -pie -latomic

TNN

TNN is an open-source deep learning reasoning framework developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterTNN 0.3Target: CPU - Model: DenseNetLLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.010002000300040005000SE +/- 6.75, N = 3SE +/- 1.92, N = 3SE +/- 50.48, N = 3SE +/- 0.57, N = 32499.823634.334513.712455.64-fopenmp=libomp - MIN: 2420.99 / MAX: 2590.77-fopenmp=libomp - MIN: 3566.47 / MAX: 3716.42-fopenmp - MIN: 4338.8 / MAX: 4643.28-fopenmp=libomp - MIN: 2405.98 / MAX: 2515.951. (CXX) g++ options: -O3 -march=native -flto -pthread -fvisibility=hidden -fvisibility=default -rdynamic -ldl

Target: CPU - Model: DenseNet

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./tnn: 3: ./test/TNNTest: not found

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: GPT-2 - Device: CPU - Executor: StandardLLVM Clang 14GCC 12.2AOCC 4.02K4K6K8K10KSE +/- 123.19, N = 3SE +/- 107.39, N = 12SE +/- 98.49, N = 5898089409147-flto=thin-flto=auto -fno-fat-lto-objects-flto=thin1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt

Model: GPT-2 - Device: CPU - Executor: Standard

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: yolov4 - Device: CPU - Executor: StandardLLVM Clang 14GCC 12.2AOCC 4.0120240360480600SE +/- 5.06, N = 3SE +/- 21.40, N = 9SE +/- 3.33, N = 3566564576-flto=thin-flto=auto -fno-fat-lto-objects-flto=thin1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt

Model: yolov4 - Device: CPU - Executor: Standard

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.0Algorithm: SHA256LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.09000M18000M27000M36000M45000MSE +/- 115974592.12, N = 3SE +/- 97264370.12, N = 3SE +/- 51284060.91, N = 3SE +/- 115194474.21, N = 3SE +/- 138604481.51, N = 33791626787039948818857373853430833686423280040426163787-Qunused-arguments-Qunused-arguments-Qunused-arguments1. (CC) gcc options: -pthread -m64 -O3 -march=native -flto -lssl -lcrypto -ldl

CppPerformanceBenchmarks

CppPerformanceBenchmarks is a set of C++ compiler performance benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Math LibraryLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.04080120160200SE +/- 0.74, N = 3SE +/- 0.38, N = 3SE +/- 0.41, N = 3SE +/- 0.39, N = 3SE +/- 0.09, N = 3165.55164.39169.48166.58164.801. (CXX) g++ options: -O3 -march=native -flto -std=c++11

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: super-resolution-10 - Device: CPU - Executor: ParallelLLVM Clang 14GCC 12.2AOCC 4.02K4K6K8K10KSE +/- 110.94, N = 3SE +/- 88.15, N = 6SE +/- 69.59, N = 3894993959113-flto=thin-flto=auto -fno-fat-lto-objects-flto=thin1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt

Model: super-resolution-10 - Device: CPU - Executor: Parallel

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: GPT-2 - Device: CPU - Executor: ParallelLLVM Clang 14GCC 12.2AOCC 4.015003000450060007500SE +/- 21.71, N = 3SE +/- 34.87, N = 3SE +/- 19.11, N = 3706271267135-flto=thin-flto=auto -fno-fat-lto-objects-flto=thin1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt

Model: GPT-2 - Device: CPU - Executor: Parallel

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: bertsquad-12 - Device: CPU - Executor: ParallelLLVM Clang 14GCC 12.2AOCC 4.02004006008001000SE +/- 2.52, N = 3SE +/- 1.09, N = 3SE +/- 3.50, N = 3923918917-flto=thin-flto=auto -fno-fat-lto-objects-flto=thin1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt

Model: bertsquad-12 - Device: CPU - Executor: Parallel

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: yolov4 - Device: CPU - Executor: ParallelLLVM Clang 14GCC 12.2AOCC 4.0120240360480600SE +/- 0.44, N = 3SE +/- 0.29, N = 3SE +/- 0.73, N = 3563568569-flto=thin-flto=auto -fno-fat-lto-objects-flto=thin1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt

Model: yolov4 - Device: CPU - Executor: Parallel

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: ArcFace ResNet-100 - Device: CPU - Executor: ParallelLLVM Clang 14GCC 12.2AOCC 4.05001000150020002500SE +/- 2.92, N = 3SE +/- 3.09, N = 3SE +/- 4.18, N = 3235323612356-flto=thin-flto=auto -fno-fat-lto-objects-flto=thin1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt

Model: ArcFace ResNet-100 - Device: CPU - Executor: Parallel

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

simdjson

This is a benchmark of SIMDJSON, a high performance JSON parser. SIMDJSON aims to be the fastest JSON parser and is used by projects like Microsoft FishStore, Yandex ClickHouse, Shopify, and others. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 2.0Throughput Test: KostyaLLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0246810SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 35.626.035.895.851. (CXX) g++ options: -O3 -march=native -flto

Throughput Test: Kostya

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./simdjson: 3: ./bench_ondemand: not found

ONNX Runtime

ONNX Runtime is developed by Microsoft and partners as a open-source, cross-platform, high performance machine learning inferencing and training accelerator. This test profile runs the ONNX Runtime with various models available from the ONNX Zoo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.11Model: fcn-resnet101-11 - Device: CPU - Executor: ParallelLLVM Clang 14GCC 12.2AOCC 4.0306090120150SE +/- 0.50, N = 3SE +/- 0.17, N = 3SE +/- 0.58, N = 3135134135-flto=thin-flto=auto -fno-fat-lto-objects-flto=thin1. (CXX) g++ options: -O3 -march=native -flto -ffunction-sections -fdata-sections -mtune=native -ldl -lrt

Model: fcn-resnet101-11 - Device: CPU - Executor: Parallel

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./onnx: line 2: ./onnxruntime/build/Linux/Release/onnxruntime_perf_test: No such file or directory

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 1 - Input: Bosphorus 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0246810SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 36.636.646.636.646.681. (CC) gcc options: -O3 -march=native -flto -fPIE -fPIC -O2 -pie -rdynamic -lpthread -lrt

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: regnety_400mLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.03691215SE +/- 0.12, N = 5SE +/- 0.09, N = 6SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.13, N = 39.629.4111.8412.017.55MIN: 9.1 / MAX: 10.59MIN: 9.11 / MAX: 27.31-lgomp - MIN: 11.71 / MAX: 18.24-lgomp - MIN: 11.85 / MAX: 18.93-lomp - MIN: 7.32 / MAX: 8.831. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: squeezenet_ssdLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.03691215SE +/- 0.11, N = 5SE +/- 0.10, N = 6SE +/- 0.02, N = 3SE +/- 0.09, N = 3SE +/- 0.21, N = 311.9711.8012.2712.4311.27MIN: 11.52 / MAX: 21.23MIN: 11.31 / MAX: 28.92-lgomp - MIN: 12.11 / MAX: 13.52-lgomp - MIN: 12.2 / MAX: 13.43-lomp - MIN: 10.85 / MAX: 12.531. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: yolov4-tinyLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.048121620SE +/- 0.22, N = 5SE +/- 0.08, N = 6SE +/- 0.07, N = 3SE +/- 0.06, N = 3SE +/- 0.06, N = 314.9414.6814.1214.0314.38MIN: 14.52 / MAX: 17.6MIN: 14.35 / MAX: 24.68-lgomp - MIN: 13.91 / MAX: 15.48-lgomp - MIN: 13.79 / MAX: 15.31-lomp - MIN: 14.16 / MAX: 15.321. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: resnet50LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.03691215SE +/- 0.24, N = 5SE +/- 0.18, N = 6SE +/- 0.23, N = 3SE +/- 0.19, N = 3SE +/- 0.11, N = 312.3212.3212.2612.2511.62MIN: 11.77 / MAX: 14.07MIN: 11.77 / MAX: 22.76-lgomp - MIN: 11.85 / MAX: 14.07-lgomp - MIN: 11.82 / MAX: 13.43-lomp - MIN: 11.29 / MAX: 13.291. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: alexnetLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01.16782.33563.50344.67125.839SE +/- 0.04, N = 5SE +/- 0.01, N = 6SE +/- 0.13, N = 3SE +/- 0.15, N = 3SE +/- 0.08, N = 35.165.194.984.994.79MIN: 4.97 / MAX: 7.15MIN: 5.05 / MAX: 6.8-lgomp - MIN: 4.76 / MAX: 5.88-lgomp - MIN: 4.76 / MAX: 12.84-lomp - MIN: 4.59 / MAX: 5.881. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: resnet18LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0246810SE +/- 0.20, N = 5SE +/- 0.08, N = 6SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 37.417.367.497.497.18MIN: 7.08 / MAX: 10.22MIN: 7.12 / MAX: 8.7-lgomp - MIN: 7.32 / MAX: 8.85-lgomp - MIN: 7.3 / MAX: 8.43-lomp - MIN: 6.98 / MAX: 8.341. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: vgg16LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0612182430SE +/- 0.34, N = 5SE +/- 0.42, N = 6SE +/- 0.36, N = 3SE +/- 0.26, N = 3SE +/- 0.31, N = 325.5225.9624.3924.5624.83MIN: 24.44 / MAX: 33.71MIN: 24.5 / MAX: 34.87-lgomp - MIN: 23.8 / MAX: 30.07-lgomp - MIN: 24.02 / MAX: 31.29-lomp - MIN: 24.13 / MAX: 31.131. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: googlenetLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0246810SE +/- 0.15, N = 5SE +/- 0.08, N = 6SE +/- 0.14, N = 3SE +/- 0.13, N = 3SE +/- 0.15, N = 38.638.548.658.667.85MIN: 8.24 / MAX: 9.92MIN: 8.23 / MAX: 15.46-lgomp - MIN: 8.39 / MAX: 10.25-lgomp - MIN: 8.41 / MAX: 9.8-lomp - MIN: 7.47 / MAX: 9.411. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: blazefaceLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.35780.71561.07341.43121.789SE +/- 0.02, N = 5SE +/- 0.01, N = 6SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 31.411.371.581.591.21MIN: 1.33 / MAX: 1.84MIN: 1.32 / MAX: 2.09-lgomp - MIN: 1.55 / MAX: 2.27-lgomp - MIN: 1.55 / MAX: 2.26-lomp - MIN: 1.18 / MAX: 1.941. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: efficientnet-b0LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01.0442.0883.1324.1765.22SE +/- 0.05, N = 5SE +/- 0.03, N = 6SE +/- 0.07, N = 3SE +/- 0.11, N = 3SE +/- 0.07, N = 34.354.254.574.643.85MIN: 4.17 / MAX: 17.27MIN: 4.09 / MAX: 5.43-lgomp - MIN: 4.45 / MAX: 5.39-lgomp - MIN: 4.46 / MAX: 16.81-lomp - MIN: 3.73 / MAX: 4.71. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: mnasnetLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.7471.4942.2412.9883.735SE +/- 0.04, N = 5SE +/- 0.03, N = 6SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 33.073.023.323.302.72MIN: 2.98 / MAX: 3.82MIN: 2.88 / MAX: 3.77-lgomp - MIN: 3.26 / MAX: 4.03-lgomp - MIN: 3.26 / MAX: 3.92-lomp - MIN: 2.65 / MAX: 5.371. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: shufflenet-v2LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.8461.6922.5383.3844.23SE +/- 0.03, N = 5SE +/- 0.05, N = 6SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.07, N = 33.313.253.743.762.96MIN: 3.2 / MAX: 4.88MIN: 3.05 / MAX: 3.95-lgomp - MIN: 3.68 / MAX: 12.3-lgomp - MIN: 3.69 / MAX: 4.45-lomp - MIN: 2.81 / MAX: 3.551. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU-v3-v3 - Model: mobilenet-v3LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.73351.4672.20052.9343.6675SE +/- 0.05, N = 5SE +/- 0.02, N = 6SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 32.992.993.263.242.69MIN: 2.85 / MAX: 4.22MIN: 2.9 / MAX: 3.96-lgomp - MIN: 3.19 / MAX: 3.89-lgomp - MIN: 3.18 / MAX: 4.64-lomp - MIN: 2.6 / MAX: 3.341. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU-v2-v2 - Model: mobilenet-v2LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.8371.6742.5113.3484.185SE +/- 0.03, N = 5SE +/- 0.02, N = 6SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.07, N = 33.533.443.723.713.17MIN: 3.4 / MAX: 4.36MIN: 3.31 / MAX: 4.22-lgomp - MIN: 3.64 / MAX: 4.37-lgomp - MIN: 3.64 / MAX: 5.07-lomp - MIN: 3.04 / MAX: 4.891. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: mobilenetLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.03691215SE +/- 0.10, N = 5SE +/- 0.09, N = 6SE +/- 0.11, N = 3SE +/- 0.08, N = 3SE +/- 0.11, N = 39.249.098.688.688.57MIN: 8.92 / MAX: 30.18MIN: 8.74 / MAX: 10.31-lgomp - MIN: 8.49 / MAX: 11.83-lgomp - MIN: 8.53 / MAX: 9.96-lomp - MIN: 8.29 / MAX: 9.541. (CXX) g++ options: -O3 -march=native -flto -rdynamic -lpthread

Caffe

This is a benchmark of the Caffe deep learning framework and currently supports the AlexNet and Googlenet model and execution on both CPUs and NVIDIA GPUs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: GoogleNet - Acceleration: CPU - Iterations: 100LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.020K40K60K80K100KSE +/- 326.67, N = 3SE +/- 77.99, N = 3SE +/- 221.71, N = 3SE +/- 206.69, N = 3SE +/- 160.54, N = 31074141049426498765742739341. (CXX) g++ options: -O3 -march=native -flto -fPIC -rdynamic -lglog -lgflags -lprotobuf -lcrypto -lcurl -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

simdjson

This is a benchmark of SIMDJSON, a high performance JSON parser. SIMDJSON aims to be the fastest JSON parser and is used by projects like Microsoft FishStore, Yandex ClickHouse, Shopify, and others. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 2.0Throughput Test: LargeRandomLLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.41180.82361.23541.64722.059SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 31.561.801.831.591. (CXX) g++ options: -O3 -march=native -flto

Throughput Test: LargeRandom

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./simdjson: 3: ./bench_ondemand: not found

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.02004006008001000SE +/- 0.82, N = 3SE +/- 0.92, N = 3SE +/- 1.14, N = 3SE +/- 2.28, N = 3SE +/- 0.74, N = 31137.491138.261139.641140.341138.83-fopenmp=libomp - MIN: 1132.69-fopenmp=libomp - MIN: 1134.02-fopenmp -lpthread - MIN: 1133.46-fopenmp - MIN: 1132.99-fopenmp=libomp - MIN: 1134.321. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.02004006008001000SE +/- 1.59, N = 3SE +/- 1.57, N = 3SE +/- 0.77, N = 3SE +/- 0.51, N = 3SE +/- 1.07, N = 31142.001142.951142.221139.551139.44-fopenmp=libomp - MIN: 1135.13-fopenmp=libomp - MIN: 1136.51-fopenmp -lpthread - MIN: 1135.79-fopenmp - MIN: 1134.18-fopenmp=libomp - MIN: 1134.211. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0130260390520650SE +/- 0.79, N = 3SE +/- 0.11, N = 3SE +/- 2.68, N = 3SE +/- 1.50, N = 3SE +/- 0.14, N = 3576.22575.94578.96578.82574.91-fopenmp=libomp - MIN: 572.91-fopenmp=libomp - MIN: 573.26-fopenmp -lpthread - MIN: 571.53-fopenmp - MIN: 572.91-fopenmp=libomp - MIN: 572.071. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0130260390520650SE +/- 0.73, N = 3SE +/- 0.44, N = 3SE +/- 1.15, N = 3SE +/- 2.60, N = 3SE +/- 0.21, N = 3575.13575.60582.00580.33574.01-fopenmp=libomp - MIN: 571.75-fopenmp=libomp - MIN: 571.93-fopenmp -lpthread - MIN: 575.92-fopenmp - MIN: 573.07-fopenmp=libomp - MIN: 571.161. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

JPEG XL libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.7Input: JPEG - Quality: 90LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.048121620SE +/- 0.11, N = 3SE +/- 0.11, N = 3SE +/- 0.09, N = 3SE +/- 0.04, N = 3SE +/- 0.08, N = 317.4613.1012.7612.5413.13-Xclang -mrelax-all-Xclang -mrelax-all-Xclang -mrelax-all1. (CXX) g++ options: -O3 -march=native -flto -fno-rtti -funwind-tables -O2 -fPIE -pie -latomic

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.7Input: PNG - Quality: 90LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.048121620SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.04, N = 317.9813.1913.0112.9813.50-Xclang -mrelax-all-Xclang -mrelax-all-Xclang -mrelax-all1. (CXX) g++ options: -O3 -march=native -flto -fno-rtti -funwind-tables -O2 -fPIE -pie -latomic

Monte Carlo Simulations of Ionised Nebulae

Mocassin is the Monte Carlo Simulations of Ionised Nebulae. MOCASSIN is a fully 3D or 2D photoionisation and dust radiative transfer code which employs a Monte Carlo approach to the transfer of radiation through media of arbitrary geometry and density distribution. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMonte Carlo Simulations of Ionised Nebulae 2019-03-24Input: Dust 2D tau100.0LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01530456075SE +/- 0.88, N = 3SE +/- 0.71, N = 4SE +/- 0.33, N = 3SE +/- 0.88, N = 3SE +/- 0.58, N = 364656364641. (F9X) gfortran options: -cpp -Jsource/ -ffree-line-length-0 -lm -std=legacy -O3 -O2 -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lz

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.11Encoder Speed: 0LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01530456075SE +/- 0.23, N = 3SE +/- 0.45, N = 3SE +/- 0.17, N = 3SE +/- 0.11, N = 3SE +/- 0.18, N = 365.1565.2366.2768.5263.111. (CXX) g++ options: -O3 -fPIC -march=native -flto -lm

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Person Vehicle Bike Detection FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.01.09132.18263.27394.36525.4565SE +/- 0.05, N = 4SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 34.854.834.824.80MIN: 3.25 / MAX: 15.02MIN: 3.7 / MAX: 13.95-fno-strict-overflow -fwrapv - MIN: 3.66 / MAX: 14.3MIN: 3.12 / MAX: 13.531. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Person Vehicle Bike Detection FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.0400800120016002000SE +/- 17.91, N = 4SE +/- 4.79, N = 3SE +/- 6.49, N = 3SE +/- 6.39, N = 31645.571654.661658.301664.77-fno-strict-overflow -fwrapv1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

simdjson

This is a benchmark of SIMDJSON, a high performance JSON parser. SIMDJSON aims to be the fastest JSON parser and is used by projects like Microsoft FishStore, Yandex ClickHouse, Shopify, and others. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 2.0Throughput Test: DistinctUserIDLLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.03691215SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.10, N = 310.779.289.8011.971. (CXX) g++ options: -O3 -march=native -flto

Throughput Test: DistinctUserID

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./simdjson: 3: ./bench_ondemand: not found

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 2.0Throughput Test: PartialTweetsLLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.03691215SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.06, N = 311.159.639.8511.821. (CXX) g++ options: -O3 -march=native -flto

Throughput Test: PartialTweets

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./simdjson: 3: ./bench_ondemand: not found

OpenBenchmarking.orgGB/s, More Is Bettersimdjson 2.0Throughput Test: TopTweetLLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.03691215SE +/- 0.02, N = 3SE +/- 0.09, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 310.6010.229.9010.851. (CXX) g++ options: -O3 -march=native -flto

Throughput Test: TopTweet

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./simdjson: 3: ./bench_ondemand: not found

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Person Detection FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.02004006008001000SE +/- 4.13, N = 3SE +/- 0.94, N = 3SE +/- 1.87, N = 3SE +/- 2.64, N = 31065.311066.071060.871055.29MIN: 636.05 / MAX: 1337.86MIN: 718.92 / MAX: 1326.23-fno-strict-overflow -fwrapv - MIN: 672.13 / MAX: 1319.18MIN: 875.98 / MAX: 1296.481. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Person Detection FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.0246810SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 37.477.467.507.53-fno-strict-overflow -fwrapv1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Person Detection FP32 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.02004006008001000SE +/- 5.64, N = 3SE +/- 5.67, N = 3SE +/- 7.82, N = 3SE +/- 3.44, N = 31071.501092.871085.331081.01MIN: 726.76 / MAX: 1318.28MIN: 753 / MAX: 1274.14-fno-strict-overflow -fwrapv - MIN: 731.35 / MAX: 1302.96MIN: 580.39 / MAX: 1284.591. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Person Detection FP32 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.0246810SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 37.437.297.347.37-fno-strict-overflow -fwrapv1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Face Detection FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.0120240360480600SE +/- 2.14, N = 3SE +/- 2.19, N = 3SE +/- 1.81, N = 3SE +/- 1.12, N = 3555.05555.00557.36557.01MIN: 522.63 / MAX: 589.63MIN: 537.69 / MAX: 616.06-fno-strict-overflow -fwrapv - MIN: 527.31 / MAX: 601MIN: 283.75 / MAX: 603.221. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Face Detection FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.048121620SE +/- 0.05, N = 3SE +/- 0.06, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 314.3514.3714.2914.29-fno-strict-overflow -fwrapv1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Face Detection FP16-INT8 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.060120180240300SE +/- 1.08, N = 3SE +/- 0.93, N = 3SE +/- 0.11, N = 3SE +/- 0.22, N = 3282.59283.72284.35283.07MIN: 216.72 / MAX: 375.59MIN: 146.73 / MAX: 325.51-fno-strict-overflow -fwrapv - MIN: 250.18 / MAX: 318.64MIN: 145.95 / MAX: 325.141. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Face Detection FP16-INT8 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.0714212835SE +/- 0.11, N = 3SE +/- 0.10, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 328.2728.1528.0928.19-fno-strict-overflow -fwrapv1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Machine Translation EN To DE FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.01326395265SE +/- 0.27, N = 3SE +/- 0.34, N = 3SE +/- 0.64, N = 3SE +/- 0.40, N = 359.2059.8359.4759.95MIN: 30.14 / MAX: 87.47MIN: 27.32 / MAX: 85.28-fno-strict-overflow -fwrapv - MIN: 27.89 / MAX: 86.56MIN: 28.33 / MAX: 87.441. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Machine Translation EN To DE FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.0306090120150SE +/- 0.62, N = 3SE +/- 0.76, N = 3SE +/- 1.44, N = 3SE +/- 0.90, N = 3135.00133.59134.42133.33-fno-strict-overflow -fwrapv1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

Ngspice

Ngspice is an open-source SPICE circuit simulator. Ngspice was originally based on the Berkeley SPICE electronic circuit simulator. Ngspice supports basic threading using OpenMP. This test profile is making use of the ISCAS 85 benchmark circuits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterNgspice 34Circuit: C2670LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01428425670SE +/- 0.46, N = 3SE +/- 0.14, N = 3SE +/- 0.41, N = 3SE +/- 0.14, N = 3SE +/- 0.47, N = 358.9657.7162.6863.7558.26-lstdc++-lstdc++-lstdc++-lstdc++1. (CC) gcc options: -O3 -march=native -flto -fopenmp -lm -lfftw3 -lXaw -lXmu -lXt -lXext -lX11 -lSM -lICE

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.00.05630.11260.16890.22520.2815SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.250.250.250.25MIN: 0.15 / MAX: 21.4MIN: 0.15 / MAX: 8.4-fno-strict-overflow -fwrapv - MIN: 0.15 / MAX: 8.94MIN: 0.15 / MAX: 7.461. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.014K28K42K56K70KSE +/- 31.56, N = 3SE +/- 21.27, N = 3SE +/- 73.58, N = 3SE +/- 68.00, N = 363386.5763331.0463460.6063880.92-fno-strict-overflow -fwrapv1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Age Gender Recognition Retail 0013 FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.00.0810.1620.2430.3240.405SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.360.360.360.36MIN: 0.21 / MAX: 8.85MIN: 0.21 / MAX: 8.45-fno-strict-overflow -fwrapv - MIN: 0.21 / MAX: 9.03MIN: 0.21 / MAX: 7.721. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Age Gender Recognition Retail 0013 FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.010K20K30K40K50KSE +/- 16.19, N = 3SE +/- 43.19, N = 3SE +/- 46.50, N = 3SE +/- 34.34, N = 344075.7544085.2544470.5944290.73-fno-strict-overflow -fwrapv1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Vehicle Detection FP16-INT8 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.00.97881.95762.93643.91524.894SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 34.334.324.354.32MIN: 2.65 / MAX: 13.04MIN: 2.61 / MAX: 13.39-fno-strict-overflow -fwrapv - MIN: 2.69 / MAX: 14.15MIN: 2.62 / MAX: 13.231. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Vehicle Detection FP16-INT8 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.0400800120016002000SE +/- 4.72, N = 3SE +/- 1.75, N = 3SE +/- 0.70, N = 3SE +/- 1.21, N = 31845.311851.091837.191847.73-fno-strict-overflow -fwrapv1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Vehicle Detection FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.03691215SE +/- 0.04, N = 3SE +/- 0.06, N = 3SE +/- 0.02, N = 3SE +/- 0.09, N = 310.7610.6210.5510.62MIN: 5.08 / MAX: 25.94MIN: 3.86 / MAX: 26.67-fno-strict-overflow -fwrapv - MIN: 4.88 / MAX: 26.32MIN: 4 / MAX: 26.81. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Vehicle Detection FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.0160320480640800SE +/- 2.68, N = 3SE +/- 4.19, N = 3SE +/- 1.28, N = 3SE +/- 6.50, N = 3742.44752.41757.51752.27-fno-strict-overflow -fwrapv1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Weld Porosity Detection FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.01.25782.51563.77345.03126.289SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 35.565.575.595.55MIN: 2.88 / MAX: 14.86MIN: 2.89 / MAX: 15.21-fno-strict-overflow -fwrapv - MIN: 2.91 / MAX: 14.45MIN: 2.87 / MAX: 13.91. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Weld Porosity Detection FP16 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.030060090012001500SE +/- 0.50, N = 3SE +/- 1.62, N = 3SE +/- 1.86, N = 3SE +/- 0.72, N = 31438.371434.101429.511439.92-fno-strict-overflow -fwrapv1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2022.2.devModel: Weld Porosity Detection FP16-INT8 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.01.24882.49763.74644.99526.244SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 35.525.545.555.52MIN: 2.86 / MAX: 14.5MIN: 2.86 / MAX: 13.37-fno-strict-overflow -fwrapv - MIN: 2.9 / MAX: 18.91MIN: 2.93 / MAX: 12.861. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2022.2.devModel: Weld Porosity Detection FP16-INT8 - Device: CPULLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.06001200180024003000SE +/- 1.94, N = 3SE +/- 0.54, N = 3SE +/- 2.62, N = 3SE +/- 0.61, N = 32895.022887.982879.682896.90-fno-strict-overflow -fwrapv1. (CXX) g++ options: -fPIC -O3 -march=native -flto -fsigned-char -ffunction-sections -fdata-sections -shared

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample 6000x4000 pixel JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.38Operation: SharpenLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0100200300400500SE +/- 0.88, N = 3SE +/- 2.73, N = 3SE +/- 0.67, N = 3SE +/- 0.58, N = 3SE +/- 0.67, N = 34544663413434701. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.38Operation: Noise-GaussianLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0140280420560700SE +/- 1.20, N = 3SE +/- 2.03, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.67, N = 35835876586646071. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.38Operation: ResizingLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.06001200180024003000SE +/- 4.67, N = 3SE +/- 3.79, N = 3SE +/- 8.67, N = 3SE +/- 2.52, N = 3SE +/- 5.17, N = 3265527052625274126091. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.38Operation: EnhancedLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0140280420560700SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 35835826276396121. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.38Operation: RotateLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.02004006008001000SE +/- 0.88, N = 3SE +/- 7.17, N = 3SE +/- 2.89, N = 3SE +/- 3.18, N = 3SE +/- 1.20, N = 38768969129889521. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgverify/s, More Is BetterOpenSSL 3.0Algorithm: RSA4096LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.080K160K240K320K400KSE +/- 41.04, N = 3SE +/- 38.98, N = 3SE +/- 44.62, N = 3SE +/- 54.65, N = 3SE +/- 12.58, N = 3395636.4395258.1394339.8393865.8394139.1-Qunused-arguments-Qunused-arguments-Qunused-arguments1. (CC) gcc options: -pthread -m64 -O3 -march=native -flto -lssl -lcrypto -ldl

OpenBenchmarking.orgsign/s, More Is BetterOpenSSL 3.0Algorithm: RSA4096LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.013002600390052006500SE +/- 0.55, N = 3SE +/- 10.70, N = 3SE +/- 0.27, N = 3SE +/- 1.50, N = 3SE +/- 4.01, N = 36030.96016.66029.16014.66026.2-Qunused-arguments-Qunused-arguments-Qunused-arguments1. (CC) gcc options: -pthread -m64 -O3 -march=native -flto -lssl -lcrypto -ldl

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample 6000x4000 pixel JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.38Operation: SwirlLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.030060090012001500SE +/- 2.33, N = 3SE +/- 4.91, N = 3SE +/- 2.19, N = 3SE +/- 4.18, N = 3SE +/- 2.96, N = 3139013821454147314131. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.38Operation: HWB Color SpaceLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0400800120016002000SE +/- 5.36, N = 3SE +/- 9.17, N = 3SE +/- 4.93, N = 3SE +/- 1.33, N = 3SE +/- 12.35, N = 3158415801438153317101. (CC) gcc options: -fopenmp -O3 -march=native -flto -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread

Ngspice

Ngspice is an open-source SPICE circuit simulator. Ngspice was originally based on the Berkeley SPICE electronic circuit simulator. Ngspice supports basic threading using OpenMP. This test profile is making use of the ISCAS 85 benchmark circuits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterNgspice 34Circuit: C7552LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01326395265SE +/- 0.17, N = 3SE +/- 0.06, N = 3SE +/- 0.09, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 356.4856.2260.2958.2856.31-lstdc++-lstdc++-lstdc++-lstdc++1. (CC) gcc options: -O3 -march=native -flto -fopenmp -lm -lfftw3 -lXaw -lXmu -lXt -lXext -lX11 -lSM -lICE

Tachyon

This is a test of the threaded Tachyon, a parallel ray-tracing system, measuring the time to ray-trace a sample scene. The sample scene used is the Teapot scene ray-traced to 8K x 8K with 32 samples. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTachyon 0.99.2Total TimeLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01224364860SE +/- 0.16, N = 3SE +/- 0.19, N = 3SE +/- 0.20, N = 3SE +/- 0.22, N = 3SE +/- 0.19, N = 355.0854.9755.2855.4855.491. (CC) gcc options: -m64 -O3 -fomit-frame-pointer -ffast-math -ltachyon -lm -lpthread

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.2Encoder Mode: Preset 4 - Input: Bosphorus 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.69951.3992.09852.7983.4975SE +/- 0.012, N = 3SE +/- 0.008, N = 3SE +/- 0.012, N = 3SE +/- 0.008, N = 3SE +/- 0.006, N = 33.0583.0203.1093.0673.0781. (CXX) g++ options: -O3 -march=native -flto -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Redis

Redis is an open-source in-memory data structure store, used as a database, cache, and message broker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 7.0.4Test: SET - Parallel Connections: 50GCC 13.0 14 NovGCC 12.2AOCC 4.01000K2000K3000K4000K5000KSE +/- 63650.32, N = 15SE +/- 56367.98, N = 15SE +/- 43555.06, N = 154675697.004498344.354701871.501. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3 -march=native -flto

Test: SET - Parallel Connections: 50

LLVM Clang 14: The test run did not produce a result. E: ./redis: 12: ./src/redis-benchmark: not found

LLVM Clang 15.0.2: The test run did not produce a result. E: ./redis: 12: ./src/redis-benchmark: not found

Crypto++

Crypto++ is a C++ class library of cryptographic algorithms. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/second, More Is BetterCrypto++ 8.2Test: Unkeyed AlgorithmsLLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.0130260390520650SE +/- 4.73, N = 3SE +/- 5.89, N = 3SE +/- 3.68, N = 3SE +/- 0.53, N = 3596.87593.21481.76591.181. (CXX) g++ options: -O3 -march=native -flto -fPIC -pthread -pipe

Test: Unkeyed Algorithms

GCC 13.0 14 Nov: The test quit with a non-zero exit status.

Kripke

Kripke is a simple, scalable, 3D Sn deterministic particle transport code. Its primary purpose is to research how data layout, programming paradigms and architectures effect the implementation and performance of Sn transport. Kripke is developed by LLNL. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgThroughput FoM, More Is BetterKripke 1.2.4LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.030M60M90M120M150MSE +/- 1079837.67, N = 3SE +/- 1260500.21, N = 4SE +/- 1504986.40, N = 4SE +/- 1385510.65, N = 15SE +/- 1766612.42, N = 3108977267109794900132057600133327173125045767-fopenmp=libomp-fopenmp=libomp-fopenmp-fopenmp-fopenmp=libomp1. (CXX) g++ options: -O3 -march=native -flto

PJSIP

PJSIP is a free and open source multimedia communication library written in C language implementing standard based protocols such as SIP, SDP, RTP, STUN, TURN, and ICE. It combines signaling protocol (SIP) with rich multimedia framework and NAT traversal functionality into high level API that is portable and suitable for almost any type of systems ranging from desktops, embedded systems, to mobile handsets. This test profile is making use of pjsip-perf with both the client/server on teh system. More details on the PJSIP benchmark at https://www.pjsip.org/high-performance-sip.htm Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgResponses Per Second, More Is BetterPJSIP 2.11Method: OPTIONS, StatefulGCC 13.0 14 NovGCC 12.2AOCC 4.02K4K6K8K10KSE +/- 37.24, N = 3SE +/- 17.01, N = 3SE +/- 19.14, N = 39389933093701. (CC) gcc options: -lSDL2 -lavformat -lavcodec -lswscale -lavutil -lstdc++ -lopus -lssl -lcrypto -luuid -lm -lrt -lpthread -lasound -O3 -march=native -flto

Method: OPTIONS, Stateful

LLVM Clang 14: The test quit with a non-zero exit status. E: ./pjsip: line 5: ./pjsip-apps/bin/samples/*/pjsip-perf: No such file or directory

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./pjsip: line 5: ./pjsip-apps/bin/samples/*/pjsip-perf: No such file or directory

OpenBenchmarking.orgResponses Per Second, More Is BetterPJSIP 2.11Method: INVITEGCC 13.0 14 NovGCC 12.2AOCC 4.011002200330044005500SE +/- 57.10, N = 5SE +/- 5.24, N = 3SE +/- 64.01, N = 35338527953441. (CC) gcc options: -lSDL2 -lavformat -lavcodec -lswscale -lavutil -lstdc++ -lopus -lssl -lcrypto -luuid -lm -lrt -lpthread -lasound -O3 -march=native -flto

Method: INVITE

LLVM Clang 14: The test quit with a non-zero exit status. E: ./pjsip: line 5: ./pjsip-apps/bin/samples/*/pjsip-perf: No such file or directory

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./pjsip: line 5: ./pjsip-apps/bin/samples/*/pjsip-perf: No such file or directory

Zstd Compression

This test measures the time needed to compress/decompress a sample file (a FreeBSD disk image - FreeBSD-12.2-RELEASE-amd64-memstick.img) using Zstd compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.0Compression Level: 19 - Decompression SpeedLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.011002200330044005500SE +/- 29.56, N = 3SE +/- 21.62, N = 3SE +/- 72.80, N = 3SE +/- 20.04, N = 3SE +/- 41.19, N = 34894.64913.24988.94932.05018.61. (CC) gcc options: -O3 -march=native -flto -pthread -lz -llzma

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.0Compression Level: 19 - Compression SpeedLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.020406080100SE +/- 0.52, N = 3SE +/- 0.07, N = 3SE +/- 0.50, N = 3SE +/- 0.52, N = 3SE +/- 0.27, N = 378.878.982.280.379.11. (CC) gcc options: -O3 -march=native -flto -pthread -lz -llzma

Dragonflydb

Dragonfly is an open-source database server that is a "modern Redis replacement" that aims to be the fastest memory store while being compliant with the Redis and Memcached protocols. For benchmarking Dragonfly, Memtier_benchmark is used as a NoSQL Redis/Memcache traffic generation plus benchmarking tool developed by Redis Labs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterDragonflydb 0.6Clients: 50 - Set To Get Ratio: 1:5GCC 13.0 14 NovGCC 12.2AOCC 4.01.2M2.4M3.6M4.8M6MSE +/- 37756.01, N = 3SE +/- 28018.78, N = 3SE +/- 24619.72, N = 35757620.865707956.745728724.371. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

Clients: 50 - Set To Get Ratio: 1:5

LLVM Clang 14: The test run did not produce a result. E: ./dragonflydb: 7: ./memtier_benchmark: not found

LLVM Clang 15.0.2: The test run did not produce a result. E: ./dragonflydb: 7: ./memtier_benchmark: not found

OpenBenchmarking.orgOps/sec, More Is BetterDragonflydb 0.6Clients: 50 - Set To Get Ratio: 5:1GCC 13.0 14 NovGCC 12.2AOCC 4.01.1M2.2M3.3M4.4M5.5MSE +/- 10081.64, N = 3SE +/- 11820.30, N = 3SE +/- 18465.52, N = 35271140.495275879.835242954.591. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

Clients: 50 - Set To Get Ratio: 5:1

LLVM Clang 14: The test run did not produce a result. E: ./dragonflydb: 7: ./memtier_benchmark: not found

LLVM Clang 15.0.2: The test run did not produce a result. E: ./dragonflydb: 7: ./memtier_benchmark: not found

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.5Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0510152025SE +/- 0.02, N = 3SE +/- 0.11, N = 3SE +/- 0.08, N = 3SE +/- 0.04, N = 3SE +/- 0.11, N = 317.9217.7620.7219.3220.751. (CXX) g++ options: -O3 -march=native -flto -std=c++11 -U_FORTIFY_SOURCE -lm

JPEG XL Decoding libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is suited for JPEG XL decode performance testing to PNG output file, the pts/jpexl test is for encode performance. The JPEG XL encoding/decoding is done using the libjxl codebase. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL Decoding libjxl 0.7CPU Threads: 1LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.020406080100SE +/- 0.04, N = 3SE +/- 0.16, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.16, N = 363.3262.2966.3068.4477.27

QuantLib

QuantLib is an open-source library/framework around quantitative finance for modeling, trading and risk management scenarios. QuantLib is written in C++ with Boost and its built-in benchmark used reports the QuantLib Benchmark Index benchmark score. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMFLOPS, More Is BetterQuantLib 1.21LLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.010002000300040005000SE +/- 83.36, N = 12SE +/- 66.71, N = 3SE +/- 52.72, N = 12SE +/- 7.07, N = 34783.94669.54501.14532.31. (CXX) g++ options: -O3 -march=native -rdynamic

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: unknown location(0): fatal error: in "QuantLib benchmark suite/ShortRateModel__Swaps": QuantLib::Error: cannot roll the asset back to0 (it is already at t = 1.75068)

CppPerformanceBenchmarks

CppPerformanceBenchmarks is a set of C++ compiler performance benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Stepanov VectorLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0918273645SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.06, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 340.2840.1835.5135.2039.381. (CXX) g++ options: -O3 -march=native -flto -std=c++11

Zstd Compression

This test measures the time needed to compress/decompress a sample file (a FreeBSD disk image - FreeBSD-12.2-RELEASE-amd64-memstick.img) using Zstd compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.0Compression Level: 19, Long Mode - Decompression SpeedLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.011002200330044005500SE +/- 45.42, N = 3SE +/- 33.82, N = 3SE +/- 81.06, N = 3SE +/- 49.55, N = 3SE +/- 24.94, N = 34928.95055.85073.84974.45117.81. (CC) gcc options: -O3 -march=native -flto -pthread -lz -llzma

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.0Compression Level: 19, Long Mode - Compression SpeedLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01224364860SE +/- 0.03, N = 3SE +/- 0.12, N = 3SE +/- 0.09, N = 3SE +/- 0.17, N = 3SE +/- 0.15, N = 355.354.654.354.654.61. (CC) gcc options: -O3 -march=native -flto -pthread -lz -llzma

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.16330.32660.48990.65320.8165SE +/- 0.000306, N = 3SE +/- 0.000668, N = 3SE +/- 0.016807, N = 15SE +/- 0.034180, N = 12SE +/- 0.000272, N = 30.5823810.5833770.6629050.7259780.577834-fopenmp=libomp - MIN: 0.57-fopenmp=libomp - MIN: 0.57-fopenmp -lpthread - MIN: 0.58-fopenmp - MIN: 0.58-fopenmp=libomp - MIN: 0.561. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

SQLite Speedtest

This is a benchmark of SQLite's speedtest1 benchmark program with an increased problem size of 1,000. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite Speedtest 3.30Timed Time - Size 1,000LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0816243240SE +/- 0.30, N = 3SE +/- 0.08, N = 3SE +/- 0.22, N = 3SE +/- 0.15, N = 3SE +/- 0.17, N = 336.0236.0734.7334.2334.771. (CC) gcc options: -O3 -march=native -flto -lz

Caffe

This is a benchmark of the Caffe deep learning framework and currently supports the AlexNet and Googlenet model and execution on both CPUs and NVIDIA GPUs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilli-Seconds, Fewer Is BetterCaffe 2020-02-13Model: AlexNet - Acceleration: CPU - Iterations: 100LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.010K20K30K40K50KSE +/- 113.58, N = 3SE +/- 157.80, N = 3SE +/- 57.54, N = 3SE +/- 77.62, N = 3SE +/- 59.47, N = 346800469762369623685276531. (CXX) g++ options: -O3 -march=native -flto -fPIC -rdynamic -lglog -lgflags -lprotobuf -lcrypto -lcurl -lpthread -lsz -lz -ldl -lm -llmdb -lopenblas

Zstd Compression

This test measures the time needed to compress/decompress a sample file (a FreeBSD disk image - FreeBSD-12.2-RELEASE-amd64-memstick.img) using Zstd compression with options for different compression levels / settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.0Compression Level: 3 - Decompression SpeedLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.012002400360048006000SE +/- 4.13, N = 3SE +/- 59.43, N = 3SE +/- 8.47, N = 3SE +/- 8.08, N = 3SE +/- 57.07, N = 35458.65463.55485.85432.15461.21. (CC) gcc options: -O3 -march=native -flto -pthread -lz -llzma

OpenBenchmarking.orgMB/s, More Is BetterZstd Compression 1.5.0Compression Level: 3 - Compression SpeedLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.011002200330044005500SE +/- 28.62, N = 3SE +/- 13.86, N = 3SE +/- 10.83, N = 3SE +/- 14.32, N = 3SE +/- 10.48, N = 35030.85045.24976.24978.65092.21. (CC) gcc options: -O3 -march=native -flto -pthread -lz -llzma

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.11Encoder Speed: 2LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0816243240SE +/- 0.04, N = 3SE +/- 0.12, N = 3SE +/- 0.16, N = 3SE +/- 0.24, N = 3SE +/- 0.08, N = 332.3132.3833.7534.8531.381. (CXX) g++ options: -O3 -fPIC -march=native -flto -lm

CppPerformanceBenchmarks

CppPerformanceBenchmarks is a set of C++ compiler performance benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: CtypeLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0816243240SE +/- 0.20, N = 3SE +/- 0.01, N = 3SE +/- 0.36, N = 3SE +/- 0.22, N = 3SE +/- 0.44, N = 331.7831.6333.7732.2831.361. (CXX) g++ options: -O3 -march=native -flto -std=c++11

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 8 - Buffer Length: 256 - Filter Length: 57LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0200M400M600M800M1000MSE +/- 84936093.08, N = 12SE +/- 475966.85, N = 3SE +/- 2781883.85, N = 3SE +/- 1506567.11, N = 3SE +/- 6969075.34, N = 35663527508098233336508766678345066678962433331. (CC) gcc options: -O3 -march=native -flto -pthread -lm -lc -lliquid

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: VMAF Optimized - Input: Bosphorus 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0306090120150SE +/- 1.81, N = 15SE +/- 1.13, N = 15SE +/- 1.85, N = 15SE +/- 1.87, N = 15SE +/- 1.06, N = 15110.81111.35110.49109.95112.441. (CC) gcc options: -O3 -fcommon -march=native -flto -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

WebP Image Encode

This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.2.4Encode Settings: Quality 100, Lossless, Highest CompressionLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.20250.4050.60750.811.0125SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.900.830.890.880.89-lpng16 -ljpeg -ltiff-lpng16 -ljpeg -ltiff-lpng16 -ljpeg1. (CC) gcc options: -fvisibility=hidden -O3 -march=native -flto -lm

CppPerformanceBenchmarks

CppPerformanceBenchmarks is a set of C++ compiler performance benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: AtolLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0612182430SE +/- 0.29, N = 4SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 326.0625.4625.5525.4325.521. (CXX) g++ options: -O3 -march=native -flto -std=c++11

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.1180.2360.3540.4720.59SE +/- 0.000600, N = 3SE +/- 0.000115, N = 3SE +/- 0.029304, N = 15SE +/- 0.002131, N = 3SE +/- 0.000214, N = 30.3437490.3435950.5242830.4799060.338581-fopenmp=libomp - MIN: 0.33-fopenmp=libomp - MIN: 0.33-fopenmp -lpthread - MIN: 0.35-fopenmp - MIN: 0.38-fopenmp=libomp - MIN: 0.331. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

Coremark

This is a test of EEMBC CoreMark processor benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations/Sec, More Is BetterCoremark 1.0CoreMark Size 666 - Iterations Per SecondLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0200K400K600K800K1000KSE +/- 368.05, N = 3SE +/- 521.77, N = 3SE +/- 1075.50, N = 3SE +/- 799.86, N = 3SE +/- 973.61, N = 31004815.01981244.491117817.251066150.95946032.291. (CC) gcc options: -O2 -O3 -march=native -flto -lrt" -lrt

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.0Preset: ExhaustiveLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.45620.91241.36861.82482.281SE +/- 0.0022, N = 3SE +/- 0.0029, N = 3SE +/- 0.0024, N = 3SE +/- 0.0035, N = 3SE +/- 0.0024, N = 31.98861.98142.02762.00712.01601. (CXX) g++ options: -O3 -march=native -flto -pthread

libjpeg-turbo tjbench

tjbench is a JPEG decompression/compression benchmark that is part of libjpeg-turbo, a JPEG image codec library optimized for SIMD instructions on modern CPU architectures. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMegapixels/sec, More Is Betterlibjpeg-turbo tjbench 2.1.0Test: Decompression ThroughputLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.070140210280350SE +/- 0.23, N = 3SE +/- 0.31, N = 3SE +/- 0.46, N = 3SE +/- 0.85, N = 3SE +/- 0.79, N = 3321.22320.16322.49314.76318.291. (CC) gcc options: -O3 -march=native -flto -rdynamic -lm

TNN

TNN is an open-source deep learning reasoning framework developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterTNN 0.3Target: CPU - Model: MobileNet v2LLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.090180270360450SE +/- 0.07, N = 3SE +/- 1.19, N = 3SE +/- 0.97, N = 3SE +/- 0.24, N = 3337.48431.56412.94244.05-fopenmp=libomp - MIN: 336.45 / MAX: 345.73-fopenmp=libomp - MIN: 428.32 / MAX: 434.62-fopenmp - MIN: 410.59 / MAX: 431.69-fopenmp=libomp - MIN: 243.09 / MAX: 248.961. (CXX) g++ options: -O3 -march=native -flto -pthread -fvisibility=hidden -fvisibility=default -rdynamic -ldl

Target: CPU - Model: MobileNet v2

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./tnn: 3: ./test/TNNTest: not found

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.08150.1630.24450.3260.4075SE +/- 0.001605, N = 3SE +/- 0.004019, N = 3SE +/- 0.003020, N = 15SE +/- 0.003679, N = 15SE +/- 0.002915, N = 30.3622850.3377600.3564070.3565460.347017-fopenmp=libomp - MIN: 0.32-fopenmp=libomp - MIN: 0.31-fopenmp -lpthread - MIN: 0.3-fopenmp - MIN: 0.3-fopenmp=libomp - MIN: 0.311. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

eSpeak-NG Speech Engine

This test times how long it takes the eSpeak speech synthesizer to read Project Gutenberg's The Outline of Science and output to a WAV file. This test profile is now tracking the eSpeak-NG version of eSpeak. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BettereSpeak-NG Speech Engine 20200907Text-To-Speech SynthesisLLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0510152025SE +/- 0.11, N = 4SE +/- 0.14, N = 4SE +/- 0.11, N = 4SE +/- 0.11, N = 420.5417.7717.8314.541. (CC) gcc options: -O3 -march=native -flto -std=c99

Text-To-Speech Synthesis

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./espeak: 3: ./espeak-ng: not found

JPEG XL Decoding libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is suited for JPEG XL decode performance testing to PNG output file, the pts/jpexl test is for encode performance. The JPEG XL encoding/decoding is done using the libjxl codebase. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL Decoding libjxl 0.7CPU Threads: AllLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.060120180240300SE +/- 1.97, N = 3SE +/- 1.06, N = 3SE +/- 1.42, N = 3SE +/- 3.47, N = 3SE +/- 2.69, N = 3266.89264.48245.54263.90286.44

C-Ray

This is a test of C-Ray, a simple raytracer designed to test the floating-point CPU performance. This test is multi-threaded (16 threads per core), will shoot 8 rays per pixel for anti-aliasing, and will generate a 1600 x 1200 image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 1.1Total Time - 4K, 16 Rays Per PixelLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0612182430SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.06, N = 320.1520.1319.3318.9927.451. (CC) gcc options: -lm -lpthread -O3 -march=native -flto

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.77951.5592.33853.1183.8975SE +/- 0.03887, N = 3SE +/- 0.00364, N = 3SE +/- 0.00889, N = 3SE +/- 0.00041, N = 3SE +/- 0.01463, N = 33.464443.395813.426863.364123.27367-fopenmp=libomp - MIN: 3.26-fopenmp=libomp - MIN: 3.3-fopenmp -lpthread - MIN: 3.24-fopenmp - MIN: 3.21-fopenmp=libomp - MIN: 3.131. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.09480.18960.28440.37920.474SE +/- 0.000120, N = 3SE +/- 0.000027, N = 3SE +/- 0.000894, N = 3SE +/- 0.000200, N = 3SE +/- 0.000864, N = 30.4085940.4084320.4188600.4214190.405097-fopenmp=libomp - MIN: 0.4-fopenmp=libomp - MIN: 0.39-fopenmp -lpthread - MIN: 0.4-fopenmp - MIN: 0.41-fopenmp=libomp - MIN: 0.391. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

AOBench

AOBench is a lightweight ambient occlusion renderer, written in C. The test profile is using a size of 2048 x 2048. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterAOBenchSize: 2048 x 2048 - Total TimeLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0510152025SE +/- 0.02, N = 3SE +/- 0.28, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 321.3921.7019.9019.7621.581. (CC) gcc options: -lm -O3 -march=native -flto

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 32 - Buffer Length: 256 - Filter Length: 57LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0400M800M1200M1600M2000MSE +/- 983756.97, N = 3SE +/- 721110.26, N = 3SE +/- 4147824.06, N = 3SE +/- 1289702.81, N = 3SE +/- 2450170.06, N = 3165493333317236000001633166667168060000018866000001. (CC) gcc options: -O3 -march=native -flto -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 16 - Buffer Length: 256 - Filter Length: 57LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0400M800M1200M1600M2000MSE +/- 883804.91, N = 3SE +/- 1993600.87, N = 3SE +/- 2273274.68, N = 3SE +/- 6222093.25, N = 3SE +/- 5356097.25, N = 3154426666715840666671274433333162723333317548333331. (CC) gcc options: -O3 -march=native -flto -pthread -lm -lc -lliquid

TNN

TNN is an open-source deep learning reasoning framework developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterTNN 0.3Target: CPU - Model: SqueezeNet v1.1LLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.080160240320400SE +/- 1.42, N = 3SE +/- 0.06, N = 3SE +/- 0.23, N = 3SE +/- 1.12, N = 3232.44302.02388.44217.84-fopenmp=libomp - MIN: 230.61 / MAX: 241.13-fopenmp=libomp - MIN: 301.67 / MAX: 304.24-fopenmp - MIN: 387.68 / MAX: 389.12-fopenmp=libomp - MIN: 215.42 / MAX: 219.121. (CXX) g++ options: -O3 -march=native -flto -pthread -fvisibility=hidden -fvisibility=default -rdynamic -ldl

Target: CPU - Model: SqueezeNet v1.1

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./tnn: 3: ./test/TNNTest: not found

QuadRay

VectorChief's QuadRay is a real-time ray-tracing engine written to support SIMD across ARM, MIPS, PPC, and x86/x86_64 processors. QuadRay supports SSE/SSE2/SSE4 and AVX/AVX2/AVX-512 usage on Intel/AMD CPUs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterQuadRay 2022.05.25Scene: 5 - Resolution: 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.4680.9361.4041.8722.34SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 31.922.041.932.082.051. (CXX) g++ options: -O3 -pthread -lm -lstdc++ -lX11 -lXext -lpthread

OpenBenchmarking.orgFPS, More Is BetterQuadRay 2022.05.25Scene: 1 - Resolution: 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0612182430SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.09, N = 3SE +/- 0.08, N = 3SE +/- 0.12, N = 326.1526.2326.3026.4826.431. (CXX) g++ options: -O3 -pthread -lm -lstdc++ -lX11 -lXext -lpthread

OpenBenchmarking.orgFPS, More Is BetterQuadRay 2022.05.25Scene: 5 - Resolution: 1080pLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0246810SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 37.608.077.658.228.121. (CXX) g++ options: -O3 -pthread -lm -lstdc++ -lX11 -lXext -lpthread

OpenBenchmarking.orgFPS, More Is BetterQuadRay 2022.05.25Scene: 1 - Resolution: 1080pLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.020406080100SE +/- 0.11, N = 3SE +/- 0.05, N = 3SE +/- 0.20, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3102.34101.89102.06102.31102.791. (CXX) g++ options: -O3 -pthread -lm -lstdc++ -lX11 -lXext -lpthread

CLOMP

CLOMP is the C version of the Livermore OpenMP benchmark developed to measure OpenMP overheads and other performance impacts due to threading in order to influence future system designs. This particular test profile configuration is currently set to look at the OpenMP static schedule speed-up across all available CPU cores using the recommended test configuration. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSpeedup, More Is BetterCLOMP 1.2Static OMP SpeedupLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0612182430SE +/- 0.15, N = 3SE +/- 0.23, N = 6SE +/- 0.03, N = 3SE +/- 0.27, N = 3SE +/- 0.27, N = 523.623.722.622.825.01. (CC) gcc options: -fopenmp -O3 -march=native -flto -lm

POV-Ray

This is a test of POV-Ray, the Persistence of Vision Raytracer. POV-Ray is used to create 3D graphics using ray-tracing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPOV-Ray 3.7.0.7Trace TimeLLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.048121620SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 314.8316.5715.9616.10-R/usr/lib1. (CXX) g++ options: -pipe -O3 -ffast-math -march=native -flto -lXpm -lSM -lICE -lX11 -ltiff -ljpeg -lpng -lz -lrt -lm -lboost_thread -lboost_system

Trace Time

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./povray: 3: ./unix/povray: not found

FLAC Audio Encoding

This test times how long it takes to encode a sample WAV file to FLAC audio format ten times using the --best preset settings. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterFLAC Audio Encoding 1.4WAV To FLACLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.03691215SE +/- 0.007, N = 5SE +/- 0.025, N = 5SE +/- 0.021, N = 5SE +/- 0.073, N = 5SE +/- 0.018, N = 59.4359.28010.81010.5209.9281. (CXX) g++ options: -O3 -march=native -flto -fvisibility=hidden -logg -lm

CppPerformanceBenchmarks

CppPerformanceBenchmarks is a set of C++ compiler performance benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Stepanov AbstractionLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.048121620SE +/- 0.03, N = 3SE +/- 0.10, N = 3SE +/- 0.11, N = 3SE +/- 0.14, N = 3SE +/- 0.15, N = 317.1817.0315.7315.7316.931. (CXX) g++ options: -O3 -march=native -flto -std=c++11

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.0Preset: ThoroughLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0510152025SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 319.4319.2819.7719.4519.911. (CXX) g++ options: -O3 -march=native -flto -pthread

TNN

TNN is an open-source deep learning reasoning framework developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterTNN 0.3Target: CPU - Model: SqueezeNet v2LLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.01632486480SE +/- 0.31, N = 15SE +/- 0.37, N = 15SE +/- 0.89, N = 3SE +/- 0.41, N = 1249.4361.1169.9244.59-fopenmp=libomp - MIN: 45.34 / MAX: 50.65-fopenmp=libomp - MIN: 56.49 / MAX: 63.12-fopenmp - MIN: 68.11 / MAX: 71.55-fopenmp=libomp - MIN: 40.15 / MAX: 45.731. (CXX) g++ options: -O3 -march=native -flto -pthread -fvisibility=hidden -fvisibility=default -rdynamic -ldl

Target: CPU - Model: SqueezeNet v2

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./tnn: 3: ./test/TNNTest: not found

Sockperf

This is a network socket API performance benchmark developed by Mellanox. This test profile runs both the client and server on the local host for evaluating individual system performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMessages Per Second, More Is BetterSockperf 3.7Test: ThroughputLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0200K400K600K800K1000KSE +/- 3955.76, N = 5SE +/- 4602.76, N = 5SE +/- 2684.21, N = 5SE +/- 3167.67, N = 5SE +/- 1831.17, N = 58976618837569003108891038790951. (CXX) g++ options: --param -O3 -march=native -flto -rdynamic

PJSIP

PJSIP is a free and open source multimedia communication library written in C language implementing standard based protocols such as SIP, SDP, RTP, STUN, TURN, and ICE. It combines signaling protocol (SIP) with rich multimedia framework and NAT traversal functionality into high level API that is portable and suitable for almost any type of systems ranging from desktops, embedded systems, to mobile handsets. This test profile is making use of pjsip-perf with both the client/server on teh system. More details on the PJSIP benchmark at https://www.pjsip.org/high-performance-sip.htm Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgResponses Per Second, More Is BetterPJSIP 2.11Method: OPTIONS, StatelessGCC 13.0 14 NovGCC 12.2AOCC 4.030K60K90K120K150KSE +/- 236.11, N = 3SE +/- 794.35, N = 3SE +/- 515.14, N = 31240281258251245231. (CC) gcc options: -lSDL2 -lavformat -lavcodec -lswscale -lavutil -lstdc++ -lopus -lssl -lcrypto -luuid -lm -lrt -lpthread -lasound -O3 -march=native -flto

Method: OPTIONS, Stateless

LLVM Clang 14: The test quit with a non-zero exit status. E: ./pjsip: line 5: ./pjsip-apps/bin/samples/*/pjsip-perf: No such file or directory

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./pjsip: line 5: ./pjsip-apps/bin/samples/*/pjsip-perf: No such file or directory

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.1Video Input: Bosphorus 4K - Video Preset: Very FastLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01122334455SE +/- 0.22, N = 3SE +/- 0.17, N = 3SE +/- 0.61, N = 3SE +/- 0.42, N = 3SE +/- 0.42, N = 345.3144.6244.1241.7847.33-lpthread-lpthread1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O3 -march=native -flto -lm -lrt

C-Blosc

C-Blosc (c-blosc2) simple, compressed, fast and persistent data store library for C. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterC-Blosc 2.3Test: blosclz bitshuffleLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.03K6K9K12K15KSE +/- 150.84, N = 3SE +/- 152.49, N = 3SE +/- 38.45, N = 3SE +/- 76.49, N = 3SE +/- 81.98, N = 313694.213989.213286.513017.313508.9

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.05140.10280.15420.20560.257SE +/- 0.000141, N = 3SE +/- 0.000409, N = 3SE +/- 0.000270, N = 3SE +/- 0.000273, N = 3SE +/- 0.000436, N = 30.2244490.2241010.2279180.2285140.222378-fopenmp=libomp - MIN: 0.22-fopenmp=libomp - MIN: 0.22-fopenmp -lpthread - MIN: 0.22-fopenmp - MIN: 0.21-fopenmp=libomp - MIN: 0.211. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.03160.06320.09480.12640.158SE +/- 0.000185, N = 3SE +/- 0.000110, N = 3SE +/- 0.000031, N = 3SE +/- 0.000173, N = 3SE +/- 0.000006, N = 30.1363100.1363430.1403490.1398620.133169-fopenmp=libomp - MIN: 0.13-fopenmp=libomp - MIN: 0.13-fopenmp -lpthread - MIN: 0.13-fopenmp - MIN: 0.13-fopenmp=libomp - MIN: 0.131. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.0Preset: FastLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.090180270360450SE +/- 0.42, N = 3SE +/- 0.36, N = 3SE +/- 0.15, N = 3SE +/- 1.00, N = 3SE +/- 0.03, N = 3412.21404.45294.57289.29427.331. (CXX) g++ options: -O3 -march=native -flto -pthread

KTX-Software toktx

This is a benchmark of The Khronos Group's KTX-Software library and tools. KTX-Software provides "toktx" for converting/creating in the KTX container format for image textures. This benchmark times how long it takes to convert to KTX 2.0 format with various settings using a reference PNG sample input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterKTX-Software toktx 4.0Settings: Zstd Compression 19LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.03691215SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.07, N = 3SE +/- 0.02, N = 311.4611.6411.6211.6911.35

WebP Image Encode

This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.2.4Encode Settings: Quality 100, LosslessLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.5131.0261.5392.0522.565SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 32.152.062.282.232.17-lpng16 -ljpeg -ltiff-lpng16 -ljpeg -ltiff-lpng16 -ljpeg1. (CC) gcc options: -fvisibility=hidden -O3 -march=native -flto -lm

Dolfyn

Dolfyn is a Computational Fluid Dynamics (CFD) code of modern numerical simulation techniques. The Dolfyn test profile measures the execution time of the bundled computational fluid dynamics demos that are bundled with Dolfyn. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterDolfyn 0.527Computational Fluid DynamicsLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.03691215SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.06, N = 3SE +/- 0.04, N = 310.8210.7710.5110.7910.77

OpenJPEG

OpenJPEG is an open-source JPEG 2000 codec written in the C programming language. The default input for this test profile is the NASA/JPL-Caltech/MSSS Curiosity panorama 717MB TIFF image file converting to JPEG2000 format. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOpenJPEG 2.4Encode: NASA Curiosity Panorama M34LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.016K32K48K64K80KSE +/- 79.04, N = 3SE +/- 566.51, N = 7SE +/- 357.45, N = 3SE +/- 575.75, N = 15SE +/- 115.52, N = 362689644166960871222739911. (CXX) g++ options: -O3 -march=native -flto -rdynamic

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.5Encoder Mode: Speed 10 Realtime - Input: Bosphorus 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01530456075SE +/- 0.21, N = 3SE +/- 0.19, N = 3SE +/- 0.12, N = 3SE +/- 0.03, N = 3SE +/- 0.19, N = 368.7164.8268.3068.2465.441. (CXX) g++ options: -O3 -march=native -flto -std=c++11 -U_FORTIFY_SOURCE -lm

Xsbench

XSBench is a mini-app representing a key computational kernel of the Monte Carlo neutronics application OpenMC. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgLookups/s, More Is BetterXsbench 2017-07-06LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01.5M3M4.5M6M7.5MSE +/- 2015.53, N = 3SE +/- 7500.25, N = 3SE +/- 4489.29, N = 3SE +/- 1443.86, N = 3SE +/- 4600.85, N = 3675223867560016318262623450870813461. (CC) gcc options: -std=gnu99 -fopenmp -O3 -lm

CppPerformanceBenchmarks

CppPerformanceBenchmarks is a set of C++ compiler performance benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Function ObjectsLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.03691215SE +/- 0.006, N = 3SE +/- 0.014, N = 3SE +/- 0.094, N = 3SE +/- 0.086, N = 3SE +/- 0.031, N = 39.3339.2469.2119.2519.0231. (CXX) g++ options: -O3 -march=native -flto -std=c++11

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.34740.69481.04221.38961.737SE +/- 0.01491, N = 3SE +/- 0.01741, N = 3SE +/- 0.00799, N = 3SE +/- 0.00876, N = 3SE +/- 0.00757, N = 31.513691.511761.533051.543791.50016-fopenmp=libomp - MIN: 1.45-fopenmp=libomp - MIN: 1.45-fopenmp -lpthread - MIN: 1.46-fopenmp - MIN: 1.46-fopenmp=libomp - MIN: 1.451. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.2Encoder Mode: Preset 8 - Input: Bosphorus 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.020406080100SE +/- 0.42, N = 3SE +/- 0.17, N = 3SE +/- 0.70, N = 3SE +/- 0.25, N = 3SE +/- 0.54, N = 374.8274.6176.3676.4076.511. (CXX) g++ options: -O3 -march=native -flto -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

C-Blosc

C-Blosc (c-blosc2) simple, compressed, fast and persistent data store library for C. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterC-Blosc 2.3Test: blosclz shuffleLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.05K10K15K20K25KSE +/- 173.15, N = 3SE +/- 81.78, N = 3SE +/- 203.38, N = 3SE +/- 90.86, N = 3SE +/- 238.69, N = 324197.724796.824105.023526.423837.2

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.1Video Input: Bosphorus 4K - Video Preset: Ultra FastLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.020406080100SE +/- 0.30, N = 3SE +/- 0.28, N = 3SE +/- 0.22, N = 3SE +/- 0.29, N = 3SE +/- 0.03, N = 384.8082.9782.5680.3185.35-lpthread-lpthread1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O3 -march=native -flto -lm -lrt

Google Draco

Draco is a library developed by Google for compressing/decompressing 3D geometric meshes and point clouds. This test profile uses some Artec3D PLY models as the sample 3D model input formats for Draco compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterGoogle Draco 1.5.0Model: Church FacadeLLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.010002000300040005000SE +/- 47.16, N = 6SE +/- 8.51, N = 3SE +/- 1.00, N = 3SE +/- 7.54, N = 346494782442443241. (CXX) g++ options: -O3 -march=native -flto

Model: Church Facade

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./draco: 3: ./draco_encoder: not found

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: Visual Quality Optimized - Input: Bosphorus 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0306090120150SE +/- 0.17, N = 3SE +/- 0.26, N = 3SE +/- 0.23, N = 3SE +/- 0.07, N = 3SE +/- 0.03, N = 3111.24110.25111.61111.60111.781. (CC) gcc options: -O3 -fcommon -march=native -flto -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 7 - Input: Bosphorus 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.020406080100SE +/- 0.07, N = 3SE +/- 0.35, N = 3SE +/- 0.19, N = 3SE +/- 0.16, N = 3SE +/- 0.09, N = 3105.69105.55107.16106.57108.321. (CC) gcc options: -O3 -march=native -flto -fPIE -fPIC -O2 -pie -rdynamic -lpthread -lrt

Primesieve

Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 8.0Length: 1e12LLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.0246810SE +/- 0.016, N = 3SE +/- 0.008, N = 3SE +/- 0.010, N = 3SE +/- 0.011, N = 36.2436.1726.1876.2451. (CXX) g++ options: -O3 -march=native -flto

Length: 1e12

GCC 13.0 14 Nov: The test quit with a non-zero exit status.

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01.19882.39763.59644.79525.994SE +/- 0.01647, N = 3SE +/- 0.01758, N = 3SE +/- 0.00740, N = 3SE +/- 0.00364, N = 3SE +/- 0.00469, N = 35.303375.319305.305015.327805.28785-fopenmp=libomp - MIN: 5.24-fopenmp=libomp - MIN: 5.24-fopenmp -lpthread - MIN: 5.23-fopenmp - MIN: 5.25-fopenmp=libomp - MIN: 5.231. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: PSNR/SSIM Optimized - Input: Bosphorus 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0306090120150SE +/- 0.22, N = 3SE +/- 0.15, N = 3SE +/- 0.13, N = 3SE +/- 0.11, N = 3SE +/- 0.09, N = 3121.26120.82122.36122.32121.781. (CC) gcc options: -O3 -fcommon -march=native -flto -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.0Preset: MediumLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.04080120160200SE +/- 0.11, N = 3SE +/- 0.21, N = 3SE +/- 0.51, N = 3SE +/- 0.92, N = 3SE +/- 0.27, N = 3156.45155.09135.19132.59160.961. (CXX) g++ options: -O3 -march=native -flto -pthread

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.1Video Input: Bosphorus 1080p - Video Preset: Very FastLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0306090120150SE +/- 0.27, N = 3SE +/- 0.13, N = 3SE +/- 0.32, N = 3SE +/- 0.39, N = 3SE +/- 0.04, N = 3109.33107.85108.11102.77113.72-lpthread-lpthread1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O3 -march=native -flto -lm -lrt

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.2Encoder Mode: Preset 10 - Input: Bosphorus 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0306090120150SE +/- 1.02, N = 3SE +/- 0.71, N = 3SE +/- 1.28, N = 3SE +/- 1.43, N = 3SE +/- 0.71, N = 3129.16130.11132.75131.04131.291. (CXX) g++ options: -O3 -march=native -flto -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

oneDNN

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.7Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPULLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.31670.63340.95011.26681.5835SE +/- 0.00184, N = 3SE +/- 0.00660, N = 3SE +/- 0.01473, N = 13SE +/- 0.00253, N = 3SE +/- 0.00291, N = 31.395791.407661.405031.386961.38455-fopenmp=libomp - MIN: 1.35-fopenmp=libomp - MIN: 1.35-fopenmp -lpthread - MIN: 1.34-fopenmp - MIN: 1.35-fopenmp=libomp - MIN: 1.331. (CXX) g++ options: -O3 -march=native -flto -msse4.1 -fPIC -pie -ldl

KTX-Software toktx

This is a benchmark of The Khronos Group's KTX-Software library and tools. KTX-Software provides "toktx" for converting/creating in the KTX container format for image textures. This benchmark times how long it takes to convert to KTX 2.0 format with various settings using a reference PNG sample input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterKTX-Software toktx 4.0Settings: UASTC 3LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01.12122.24243.36364.48485.606SE +/- 0.005, N = 3SE +/- 0.025, N = 3SE +/- 0.006, N = 3SE +/- 0.005, N = 3SE +/- 0.004, N = 34.9834.9824.9614.9584.957

WebP Image Encode

This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.2.4Encode Settings: Quality 100, Highest CompressionLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0246810SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 36.216.164.604.486.33-lpng16 -ljpeg -ltiff-lpng16 -ljpeg -ltiff-lpng16 -ljpeg1. (CC) gcc options: -fvisibility=hidden -O3 -march=native -flto -lm

Google Draco

Draco is a library developed by Google for compressing/decompressing 3D geometric meshes and point clouds. This test profile uses some Artec3D PLY models as the sample 3D model input formats for Draco compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterGoogle Draco 1.5.0Model: LionLLVM Clang 15.0.2LLVM Clang 14GCC 12.2AOCC 4.08001600240032004000SE +/- 22.07, N = 3SE +/- 52.92, N = 3SE +/- 30.51, N = 3SE +/- 5.51, N = 337003763348633731. (CXX) g++ options: -O3 -march=native -flto

Model: Lion

GCC 13.0 14 Nov: The test quit with a non-zero exit status. E: ./draco: 3: ./draco_encoder: not found

LAME MP3 Encoding

LAME is an MP3 encoder licensed under the LGPL. This test measures the time required to encode a WAV file to MP3 format. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterLAME MP3 Encoding 3.100WAV To MP3LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01.06472.12943.19414.25885.3235SE +/- 0.010, N = 3SE +/- 0.005, N = 3SE +/- 0.007, N = 3SE +/- 0.012, N = 3SE +/- 0.014, N = 34.5294.6204.3874.4094.732-ffast-math -funroll-loops -fschedule-insns2 -fbranch-count-reg -fforce-addr-ffast-math -funroll-loops -fschedule-insns2 -fbranch-count-reg -fforce-addr1. (CC) gcc options: -O3 -pipe -march=native -flto -lncurses -lm

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.11Encoder Speed: 6, LosslessLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.01.13132.26263.39394.52525.6565SE +/- 0.024, N = 3SE +/- 0.002, N = 3SE +/- 0.002, N = 3SE +/- 0.027, N = 3SE +/- 0.019, N = 34.3634.3774.6755.0284.2291. (CXX) g++ options: -O3 -fPIC -march=native -flto -lm

yquake2

This is a test of Yamagi Quake II. Yamagi Quake II is an enhanced client for id Software's Quake II with focus on offline and coop gameplay. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteryquake2 8.10Renderer: Software CPU Color Light - AF: Off - MSAA: Off - Resolution: 1920 x 1080GCC 13.0 14 NovGCC 12.2AOCC 4.04080120160200SE +/- 0.95, N = 3SE +/- 0.52, N = 3SE +/- 2.19, N = 4185.3176.4191.81. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC

Renderer: Software CPU Color Light - AF: Off - MSAA: Off - Resolution: 1920 x 1080

LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

LAMMPS Molecular Dynamics Simulator

LAMMPS is a classical molecular dynamics code, and an acronym for Large-scale Atomic/Molecular Massively Parallel Simulator. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 23Jun2022Model: Rhodopsin ProteinLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.048121620SE +/- 0.14, N = 15SE +/- 0.13, N = 15SE +/- 0.23, N = 3SE +/- 0.17, N = 3SE +/- 0.06, N = 317.3917.6917.0417.2617.911. (CXX) g++ options: -O3 -march=native -flto -lm -ldl

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 10 - Input: Bosphorus 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.04080120160200SE +/- 0.29, N = 3SE +/- 0.19, N = 3SE +/- 0.18, N = 3SE +/- 0.31, N = 3SE +/- 0.26, N = 3171.02170.79173.36172.53171.091. (CC) gcc options: -O3 -march=native -flto -fPIE -fPIC -O2 -pie -rdynamic -lpthread -lrt

Nettle

GNU Nettle is a low-level cryptographic library used by GnuTLS and other software. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbyte/s, More Is BetterNettle 3.8Test: aes256LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.02K4K6K8K10KSE +/- 0.32, N = 3SE +/- 47.83, N = 3SE +/- 1.22, N = 3SE +/- 53.19, N = 3SE +/- 116.25, N = 38553.158566.178490.468521.418564.64MIN: 6494.58 / MAX: 12673.87MIN: 6405.01 / MAX: 12818.56-lhogweed - MIN: 6473.53 / MAX: 12523.15-lhogweed - MIN: 6434.82 / MAX: 12801.5MIN: 6283.86 / MAX: 12924.641. (CC) gcc options: -O3 -march=native -flto -ggdb3 -lnettle -lgmp -lm -lcrypto

yquake2

This is a test of Yamagi Quake II. Yamagi Quake II is an enhanced client for id Software's Quake II with focus on offline and coop gameplay. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteryquake2 8.10Renderer: Software CPU Color Light - AF: On - MSAA: Off - Resolution: 1920 x 1080GCC 13.0 14 NovGCC 12.2AOCC 4.04080120160200SE +/- 0.50, N = 3SE +/- 0.26, N = 3SE +/- 1.77, N = 3178.5172.5188.21. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC

Renderer: Software CPU Color Light - AF: On - MSAA: Off - Resolution: 1920 x 1080

LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

OpenBenchmarking.orgFrames Per Second, More Is Betteryquake2 8.10Renderer: Software CPU Color Light - AF: On - MSAA: On - Resolution: 1920 x 1080GCC 13.0 14 NovGCC 12.2AOCC 4.04080120160200SE +/- 0.33, N = 3SE +/- 0.93, N = 3SE +/- 0.68, N = 3181.3171.2185.31. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC

Renderer: Software CPU Color Light - AF: On - MSAA: On - Resolution: 1920 x 1080

LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.2Encoder Mode: Preset 12 - Input: Bosphorus 4KLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.04080120160200SE +/- 0.49, N = 3SE +/- 0.91, N = 3SE +/- 0.64, N = 3SE +/- 0.19, N = 3SE +/- 0.95, N = 3184.15183.19188.62187.37188.401. (CXX) g++ options: -O3 -march=native -flto -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

yquake2

This is a test of Yamagi Quake II. Yamagi Quake II is an enhanced client for id Software's Quake II with focus on offline and coop gameplay. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteryquake2 8.10Renderer: Software CPU Color Light - AF: Off - MSAA: On - Resolution: 1920 x 1080GCC 13.0 14 NovGCC 12.2AOCC 4.04080120160200SE +/- 0.78, N = 3SE +/- 0.37, N = 3SE +/- 0.54, N = 3184.4176.2189.91. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC

Renderer: Software CPU Color Light - AF: Off - MSAA: On - Resolution: 1920 x 1080

LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

OpenBenchmarking.orgFrames Per Second, More Is Betteryquake2 8.10Renderer: Software CPU - AF: On - MSAA: On - Resolution: 1920 x 1080GCC 13.0 14 NovGCC 12.2AOCC 4.04080120160200SE +/- 0.43, N = 3SE +/- 0.91, N = 3SE +/- 0.15, N = 3191.4185.1196.31. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC

Renderer: Software CPU - AF: On - MSAA: On - Resolution: 1920 x 1080

LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

OpenBenchmarking.orgFrames Per Second, More Is Betteryquake2 8.10Renderer: Software CPU - AF: On - MSAA: Off - Resolution: 1920 x 1080GCC 13.0 14 NovGCC 12.2AOCC 4.04080120160200SE +/- 0.30, N = 3SE +/- 1.68, N = 3SE +/- 2.06, N = 3192.3185.9198.11. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC

Renderer: Software CPU - AF: On - MSAA: Off - Resolution: 1920 x 1080

LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

OpenBenchmarking.orgFrames Per Second, More Is Betteryquake2 8.10Renderer: Software CPU - AF: Off - MSAA: Off - Resolution: 1920 x 1080GCC 13.0 14 NovGCC 12.2AOCC 4.050100150200250SE +/- 0.63, N = 3SE +/- 0.24, N = 3SE +/- 1.48, N = 3201.0192.2205.71. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC

Renderer: Software CPU - AF: Off - MSAA: Off - Resolution: 1920 x 1080

LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.11Encoder Speed: 6LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.72051.4412.16152.8823.6025SE +/- 0.005, N = 3SE +/- 0.018, N = 3SE +/- 0.006, N = 3SE +/- 0.011, N = 3SE +/- 0.022, N = 33.0443.0793.1053.2022.9431. (CXX) g++ options: -O3 -fPIC -march=native -flto -lm

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 0.11Encoder Speed: 10, LosslessLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.00.69231.38462.07692.76923.4615SE +/- 0.018, N = 3SE +/- 0.009, N = 3SE +/- 0.004, N = 3SE +/- 0.021, N = 3SE +/- 0.033, N = 32.7252.7612.9473.0772.7731. (CXX) g++ options: -O3 -fPIC -march=native -flto -lm

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.1Video Input: Bosphorus 1080p - Video Preset: Ultra FastLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.050100150200250SE +/- 0.89, N = 3SE +/- 1.14, N = 3SE +/- 0.51, N = 3SE +/- 0.80, N = 3SE +/- 0.25, N = 3213.39210.45210.39204.74219.33-lpthread-lpthread1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O3 -march=native -flto -lm -lrt

yquake2

This is a test of Yamagi Quake II. Yamagi Quake II is an enhanced client for id Software's Quake II with focus on offline and coop gameplay. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteryquake2 8.10Renderer: Software CPU - AF: Off - MSAA: On - Resolution: 1920 x 1080GCC 13.0 14 NovGCC 12.2AOCC 4.050100150200250SE +/- 0.64, N = 3SE +/- 0.15, N = 3SE +/- 1.56, N = 3199.0192.0205.61. (CC) gcc options: -shared -lm -ldl -rdynamic -lSDL2 -O3 -march=native -flto -std=gnu99 -fno-strict-aliasing -fwrapv -fvisibility=hidden -MMD -mfpmath=sse -fPIC

Renderer: Software CPU - AF: Off - MSAA: On - Resolution: 1920 x 1080

LLVM Clang 14: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

LLVM Clang 15.0.2: The test quit with a non-zero exit status. E: ./yquake2: 5: ./quake2: not found

Nettle

GNU Nettle is a low-level cryptographic library used by GnuTLS and other software. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbyte/s, More Is BetterNettle 3.8Test: sha512LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.02004006008001000SE +/- 0.75, N = 3SE +/- 0.60, N = 3SE +/- 0.35, N = 3SE +/- 0.38, N = 3SE +/- 0.77, N = 3863.73864.87859.96858.30864.29-lhogweed-lhogweed1. (CC) gcc options: -O3 -march=native -flto -ggdb3 -lnettle -lgmp -lm -lcrypto

WebP Image Encode

This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.2.4Encode Settings: Quality 100LLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.048121620SE +/- 0.24, N = 3SE +/- 0.09, N = 3SE +/- 0.06, N = 3SE +/- 0.11, N = 3SE +/- 0.01, N = 317.0017.5116.8016.6117.67-lpng16 -ljpeg -ltiff-lpng16 -ljpeg -ltiff-lpng16 -ljpeg1. (CC) gcc options: -fvisibility=hidden -O3 -march=native -flto -lm

Nettle

GNU Nettle is a low-level cryptographic library used by GnuTLS and other software. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbyte/s, More Is BetterNettle 3.8Test: chachaLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.030060090012001500SE +/- 0.55, N = 3SE +/- 10.11, N = 3SE +/- 0.01, N = 3SE +/- 9.98, N = 3SE +/- 2.93, N = 31523.441536.841507.341515.121528.82MIN: 773.19 / MAX: 4193.52MIN: 773.42 / MAX: 4324.28-lhogweed - MIN: 757.05 / MAX: 4174.74-lhogweed - MIN: 754.38 / MAX: 4226.47MIN: 773.94 / MAX: 4229.151. (CC) gcc options: -O3 -march=native -flto -ggdb3 -lnettle -lgmp -lm -lcrypto

TSCP

This is a performance test of TSCP, Tom Kerrigan's Simple Chess Program, which has a built-in performance benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes Per Second, More Is BetterTSCP 1.81AI Chess PerformanceLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0600K1200K1800K2400K3000KSE +/- 14422.21, N = 5SE +/- 17050.36, N = 5SE +/- 3566.28, N = 5SE +/- 7641.62, N = 5SE +/- 18536.62, N = 5285702229425792492228249682124770631. (CC) gcc options: -O3 -march=native -flto

WebP Image Encode

This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.2.4Encode Settings: DefaultLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.0714212835SE +/- 0.05, N = 3SE +/- 0.40, N = 3SE +/- 0.07, N = 3SE +/- 0.09, N = 3SE +/- 0.02, N = 329.2329.4528.1828.0229.65-lpng16 -ljpeg -ltiff-lpng16 -ljpeg -ltiff-lpng16 -ljpeg1. (CC) gcc options: -fvisibility=hidden -O3 -march=native -flto -lm

Nettle

GNU Nettle is a low-level cryptographic library used by GnuTLS and other software. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbyte/s, More Is BetterNettle 3.8Test: poly1305-aesLLVM Clang 15.0.2LLVM Clang 14GCC 13.0 14 NovGCC 12.2AOCC 4.09001800270036004500SE +/- 1.66, N = 3SE +/- 4.96, N = 3SE +/- 18.65, N = 3SE +/- 6.07, N = 3SE +/- 5.55, N = 34214.744231.634188.144182.314255.91-lhogweed-lhogweed1. (CC) gcc options: -O3 -march=native -flto -ggdb3 -lnettle -lgmp -lm -lcrypto

190 Results Shown

JPEG XL libjxl
ONNX Runtime
LeelaChessZero:
  BLAS
  Eigen
Crypto++
SecureMark
JPEG XL libjxl
TNN
ONNX Runtime:
  GPT-2 - CPU - Standard
  yolov4 - CPU - Standard
OpenSSL
CppPerformanceBenchmarks
ONNX Runtime:
  super-resolution-10 - CPU - Parallel
  GPT-2 - CPU - Parallel
  bertsquad-12 - CPU - Parallel
  yolov4 - CPU - Parallel
  ArcFace ResNet-100 - CPU - Parallel
simdjson
ONNX Runtime
SVT-HEVC
NCNN:
  CPU - regnety_400m
  CPU - squeezenet_ssd
  CPU - yolov4-tiny
  CPU - resnet50
  CPU - alexnet
  CPU - resnet18
  CPU - vgg16
  CPU - googlenet
  CPU - blazeface
  CPU - efficientnet-b0
  CPU - mnasnet
  CPU - shufflenet-v2
  CPU-v3-v3 - mobilenet-v3
  CPU-v2-v2 - mobilenet-v2
  CPU - mobilenet
Caffe
simdjson
oneDNN:
  Recurrent Neural Network Training - u8s8f32 - CPU
  Recurrent Neural Network Training - bf16bf16bf16 - CPU
  Recurrent Neural Network Inference - bf16bf16bf16 - CPU
  Recurrent Neural Network Inference - u8s8f32 - CPU
JPEG XL libjxl:
  JPEG - 90
  PNG - 90
Monte Carlo Simulations of Ionised Nebulae
libavif avifenc
OpenVINO:
  Person Vehicle Bike Detection FP16 - CPU:
    ms
    FPS
simdjson:
  DistinctUserID
  PartialTweets
  TopTweet
OpenVINO:
  Person Detection FP16 - CPU:
    ms
    FPS
  Person Detection FP32 - CPU:
    ms
    FPS
  Face Detection FP16 - CPU:
    ms
    FPS
  Face Detection FP16-INT8 - CPU:
    ms
    FPS
  Machine Translation EN To DE FP16 - CPU:
    ms
    FPS
Ngspice
OpenVINO:
  Age Gender Recognition Retail 0013 FP16-INT8 - CPU:
    ms
    FPS
  Age Gender Recognition Retail 0013 FP16 - CPU:
    ms
    FPS
  Vehicle Detection FP16-INT8 - CPU:
    ms
    FPS
  Vehicle Detection FP16 - CPU:
    ms
    FPS
  Weld Porosity Detection FP16 - CPU:
    ms
    FPS
  Weld Porosity Detection FP16-INT8 - CPU:
    ms
    FPS
GraphicsMagick:
  Sharpen
  Noise-Gaussian
  Resizing
  Enhanced
  Rotate
OpenSSL:
  RSA4096:
    verify/s
    sign/s
GraphicsMagick:
  Swirl
  HWB Color Space
Ngspice
Tachyon
SVT-AV1
Redis
Crypto++
Kripke
PJSIP:
  OPTIONS, Stateful
  INVITE
Zstd Compression:
  19 - Decompression Speed
  19 - Compression Speed
Dragonflydb:
  50 - 1:5
  50 - 5:1
AOM AV1
JPEG XL Decoding libjxl
QuantLib
CppPerformanceBenchmarks
Zstd Compression:
  19, Long Mode - Decompression Speed
  19, Long Mode - Compression Speed
oneDNN
SQLite Speedtest
Caffe
Zstd Compression:
  3 - Decompression Speed
  3 - Compression Speed
libavif avifenc
CppPerformanceBenchmarks
Liquid-DSP
SVT-VP9
WebP Image Encode
CppPerformanceBenchmarks
oneDNN
Coremark
ASTC Encoder
libjpeg-turbo tjbench
TNN
oneDNN
eSpeak-NG Speech Engine
JPEG XL Decoding libjxl
C-Ray
oneDNN:
  Deconvolution Batch shapes_1d - bf16bf16bf16 - CPU
  Deconvolution Batch shapes_1d - u8s8f32 - CPU
AOBench
Liquid-DSP:
  32 - 256 - 57
  16 - 256 - 57
TNN
QuadRay:
  5 - 4K
  1 - 4K
  5 - 1080p
  1 - 1080p
CLOMP
POV-Ray
FLAC Audio Encoding
CppPerformanceBenchmarks
ASTC Encoder
TNN
Sockperf
PJSIP
Kvazaar
C-Blosc
oneDNN:
  Matrix Multiply Batch Shapes Transformer - bf16bf16bf16 - CPU
  Matrix Multiply Batch Shapes Transformer - u8s8f32 - CPU
ASTC Encoder
KTX-Software toktx
WebP Image Encode
Dolfyn
OpenJPEG
AOM AV1
Xsbench
CppPerformanceBenchmarks
oneDNN
SVT-AV1
C-Blosc
Kvazaar
Google Draco
SVT-VP9
SVT-HEVC
Primesieve
oneDNN
SVT-VP9
ASTC Encoder
Kvazaar
SVT-AV1
oneDNN
KTX-Software toktx
WebP Image Encode
Google Draco
LAME MP3 Encoding
libavif avifenc
yquake2
LAMMPS Molecular Dynamics Simulator
SVT-HEVC
Nettle
yquake2:
  Software CPU Color Light - On - Off - 1920 x 1080
  Software CPU Color Light - On - On - 1920 x 1080
SVT-AV1
yquake2:
  Software CPU Color Light - Off - On - 1920 x 1080
  Software CPU - On - On - 1920 x 1080
  Software CPU - On - Off - 1920 x 1080
  Software CPU - Off - Off - 1920 x 1080
libavif avifenc:
  6
  10, Lossless
Kvazaar
yquake2
Nettle
WebP Image Encode
Nettle
TSCP
WebP Image Encode
Nettle