dnn

qemu testing on Ubuntu 22.04 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2306107-NE-DNN37749368
Jump To Table - Results

Statistics

Remove Outliers Before Calculating Averages

Graph Settings

Prefer Vertical Bar Graphs

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Result
Identifier
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
dnn
June 10 2023
  5 Hours, 15 Minutes
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):


dnnOpenBenchmarking.orgPhoronix Test SuiteAMD Ryzen 9 7950X 16-Core (28 Cores)QEMU Standard PC (Q35 + ICH9 2009) (0.0.0 BIOS)Intel 82G33/G31/P35/P31 + ICH946GB2164GBNVIDIA GeForce RTX 4090 24GBQEMU GenericDP1080P602 x Red Hat Virtio deviceUbuntu 22.045.19.0-43-generic (x86_64)GNOME Shell 42.5X Server 1.21.1.4NVIDIA 530.41.034.6.0OpenCL 3.0 CUDA 12.1.981.3.236GCC 11.3.0 + CUDA 11.5ext41920x1080qemuProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLOpenCLVulkanCompilerFile-SystemScreen ResolutionSystem LayerDnn BenchmarksSystem Logs- Transparent Huge Pages: madvise- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-aYxV0E/gcc-11-11.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-aYxV0E/gcc-11-11.3.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v - CPU Microcode: 0xa601203- BAR1 / Visible vRAM Size: 32768 MiB - vBIOS Version: 95.02.3c.40.1a - GPU Compute Cores: 16384- Python 3.10.6- 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: disabled RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

dnnvkpeak: fp32-scalarvkpeak: fp32-vec4vkpeak: fp16-scalarvkpeak: fp16-vec4vkpeak: fp64-scalarvkpeak: fp64-vec4vkpeak: int32-scalarvkpeak: int32-vec4vkpeak: int16-scalarvkpeak: int16-vec4realsr-ncnn: 4x - Norealsr-ncnn: 4x - Yeswaifu2x-ncnn: 2x - 3 - Yesvkfft: cl-mem: Copycl-mem: Readcl-mem: Writenamd-cuda: ATPase Simulation - 327,506 Atomsvkresample: 2x - Doublevkresample: 2x - Singleoctanebench: Total Scorefahbench: clpeak: Integer Compute INTclpeak: Single-Precision Floatclpeak: Double-Precision Doubleclpeak: Global Memory Bandwidthlczero: OpenCLrodinia: OpenCL Particle Filterarrayfire: Conjugate Gradient OpenCLluxcorerender: DLSC - GPUluxcorerender: Danish Mood - GPUluxcorerender: Orange Juice - GPUluxcorerender: LuxCore Benchmark - GPUluxcorerender: Rainbow Colors and Prism - GPUfinancebench: Black-Scholes OpenCLviennacl: CPU BLAS - sCOPYviennacl: CPU BLAS - sAXPYviennacl: CPU BLAS - sDOTviennacl: CPU BLAS - dCOPYviennacl: CPU BLAS - dAXPYviennacl: CPU BLAS - dDOTviennacl: CPU BLAS - dGEMV-Nviennacl: CPU BLAS - dGEMV-Tviennacl: CPU BLAS - dGEMM-NNviennacl: CPU BLAS - dGEMM-NTviennacl: CPU BLAS - dGEMM-TNviennacl: CPU BLAS - dGEMM-TTviennacl: OpenCL BLAS - sCOPYviennacl: OpenCL BLAS - sAXPYviennacl: OpenCL BLAS - dCOPYviennacl: OpenCL BLAS - dAXPYviennacl: OpenCL BLAS - dDOTviennacl: OpenCL BLAS - dGEMV-Nviennacl: OpenCL BLAS - dGEMV-Tviennacl: OpenCL BLAS - dGEMM-NNviennacl: OpenCL BLAS - dGEMM-NTviennacl: OpenCL BLAS - dGEMM-TNviennacl: OpenCL BLAS - dGEMM-TTviennacl: OpenCL BLAS - sDOTncnn: Vulkan GPU - mobilenetncnn: Vulkan GPU-v2-v2 - mobilenet-v2ncnn: Vulkan GPU-v3-v3 - mobilenet-v3ncnn: Vulkan GPU - shufflenet-v2ncnn: Vulkan GPU - mnasnetncnn: Vulkan GPU - efficientnet-b0ncnn: Vulkan GPU - blazefacencnn: Vulkan GPU - googlenetncnn: Vulkan GPU - vgg16ncnn: Vulkan GPU - alexnetncnn: Vulkan GPU - yolov4-tinyncnn: Vulkan GPU - squeezenet_ssdncnn: Vulkan GPU - regnety_400mncnn: Vulkan GPU - vision_transformerncnn: Vulkan GPU - FastestDetncnn: Vulkan GPU - resnet18ncnn: Vulkan GPU - resnet50blender: BMW27 - NVIDIA OptiXblender: Classroom - NVIDIA OptiXblender: Fishy Cat - NVIDIA OptiXblender: Barbershop - NVIDIA OptiXblender: Pabellon Barcelona - NVIDIA OptiXindigobench: OpenCL GPU - Bedroomindigobench: OpenCL GPU - Supercarmandelgpu: GPUneatbench: GPUdnn44603.4558898.2344487.3288257.981406.301407.5344596.5944374.7629662.9939483.864.49619.7472.28199244392.9886.1801.60.0737155.2787.9111281.892068430.645840347.5579707.101396.96873.14311992.1730.928625.8319.9820.1920.8044.932.96720530929462.492.995.211112763.260.770.966.443656865176364221843511601277129713474393.031.021.291.231.011.940.761.671.681.024.493.171.50121.141.470.861.5213.477.345.5830.608.3735.50979.451830513186.54090OpenBenchmarking.org

vkpeak

Vkpeak is a Vulkan compute benchmark inspired by OpenCL's clpeak. Vkpeak provides Vulkan compute performance measurements for FP16 / FP32 / FP64 / INT16 / INT32 scalar and vec4 performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp32-scalardnn10K20K30K40K50KSE +/- 56.62, N = 344603.45

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp32-vec4dnn13K26K39K52K65KSE +/- 53.24, N = 358898.23

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp16-scalardnn10K20K30K40K50KSE +/- 39.52, N = 344487.32

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp16-vec4dnn20K40K60K80K100KSE +/- 82.01, N = 388257.98

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp64-scalardnn30060090012001500SE +/- 1.05, N = 31406.30

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp64-vec4dnn30060090012001500SE +/- 0.36, N = 31407.53

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int32-scalardnn10K20K30K40K50KSE +/- 23.36, N = 344596.59

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int32-vec4dnn10K20K30K40K50KSE +/- 25.21, N = 344374.76

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int16-scalardnn6K12K18K24K30KSE +/- 22.90, N = 329662.99

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int16-vec4dnn8K16K24K32K40KSE +/- 3.33, N = 339483.86

RealSR-NCNN

RealSR-NCNN is an NCNN neural network implementation of the RealSR project and accelerated using the Vulkan API. RealSR is the Real-World Super Resolution via Kernel Estimation and Noise Injection. NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. This test profile times how long it takes to increase the resolution of a sample image by a scale of 4x with Vulkan. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: Nodnn1.01162.02323.03484.04645.058SE +/- 0.045, N = 34.496

OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: Yesdnn510152025SE +/- 0.01, N = 319.75

Waifu2x-NCNN Vulkan

Waifu2x-NCNN is an NCNN neural network implementation of the Waifu2x converter project and accelerated using the Vulkan API. NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. This test profile times how long it takes to increase the resolution of a sample image with Vulkan. Learn more via the OpenBenchmarking.org test page.

Scale: 2x - Denoise: 3 - TAA: No

dnn: The test run did not produce a result.

OpenBenchmarking.orgSeconds, Fewer Is BetterWaifu2x-NCNN Vulkan 20200818Scale: 2x - Denoise: 3 - TAA: Yesdnn0.51321.02641.53962.05282.566SE +/- 0.024, N = 42.281

VkFFT

VkFFT is a Fast Fourier Transform (FFT) Library that is GPU accelerated by means of the Vulkan API. The VkFFT benchmark runs FFT performance differences of many different sizes before returning an overall benchmark score. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 1.1.1dnn20K40K60K80K100KSE +/- 8508.53, N = 9992441. (CXX) g++ options: -O3

Hashcat

Hashcat is an open-source, advanced password recovery tool supporting GPU acceleration with OpenCL, NVIDIA CUDA, and Radeon ROCm. Learn more via the OpenBenchmarking.org test page.

Benchmark: MD5

dnn: The test quit with a non-zero exit status. E: nvrtc: error: invalid value for --gpu-architecture (-arch)

Benchmark: SHA1

dnn: The test quit with a non-zero exit status. E: nvrtc: error: invalid value for --gpu-architecture (-arch)

Benchmark: 7-Zip

dnn: The test quit with a non-zero exit status. E: nvrtc: error: invalid value for --gpu-architecture (-arch)

Benchmark: SHA-512

dnn: The test quit with a non-zero exit status. E: nvrtc: error: invalid value for --gpu-architecture (-arch)

Benchmark: TrueCrypt RIPEMD160 + XTS

dnn: The test quit with a non-zero exit status. E: nvrtc: error: invalid value for --gpu-architecture (-arch)

Mixbench

A benchmark suite for GPUs on mixed operational intensity kernels. Learn more via the OpenBenchmarking.org test page.

Backend: OpenCL - Benchmark: Integer

dnn: The test quit with a non-zero exit status. E: ./mixbench: 3: ./mixbench-ocl-ro: not found

Backend: NVIDIA CUDA - Benchmark: Integer

dnn: The test quit with a non-zero exit status. E: ./mixbench: 3: ./mixbench-cuda-ro: not found

Backend: OpenCL - Benchmark: Double Precision

dnn: The test quit with a non-zero exit status. E: ./mixbench: 3: ./mixbench-ocl-ro: not found

Backend: OpenCL - Benchmark: Single Precision

dnn: The test quit with a non-zero exit status. E: ./mixbench: 3: ./mixbench-ocl-ro: not found

Backend: NVIDIA CUDA - Benchmark: Half Precision

dnn: The test quit with a non-zero exit status. E: ./mixbench: 3: ./mixbench-cuda-ro: not found

Backend: NVIDIA CUDA - Benchmark: Double Precision

dnn: The test quit with a non-zero exit status. E: ./mixbench: 3: ./mixbench-cuda-ro: not found

Backend: NVIDIA CUDA - Benchmark: Single Precision

dnn: The test quit with a non-zero exit status. E: ./mixbench: 3: ./mixbench-cuda-ro: not found

SHOC Scalable HeterOgeneous Computing

The CUDA and OpenCL version of Vetter's Scalable HeterOgeneous Computing benchmark suite. SHOC provides a number of different benchmark programs for evaluating the performance and stability of compute devices. Learn more via the OpenBenchmarking.org test page.

Target: OpenCL - Benchmark: S3D

dnn: The test quit with a non-zero exit status. E: ./shoc: 3: ./bin/shocdriver: not found

Target: OpenCL - Benchmark: Triad

dnn: The test quit with a non-zero exit status. E: ./shoc: 3: ./bin/shocdriver: not found

Target: OpenCL - Benchmark: FFT SP

dnn: The test quit with a non-zero exit status. E: ./shoc: 3: ./bin/shocdriver: not found

Target: OpenCL - Benchmark: MD5 Hash

dnn: The test quit with a non-zero exit status. E: ./shoc: 3: ./bin/shocdriver: not found

Target: OpenCL - Benchmark: Reduction

dnn: The test quit with a non-zero exit status. E: ./shoc: 3: ./bin/shocdriver: not found

Target: OpenCL - Benchmark: GEMM SGEMM_N

dnn: The test quit with a non-zero exit status. E: ./shoc: 3: ./bin/shocdriver: not found

Target: OpenCL - Benchmark: Max SP Flops

dnn: The test quit with a non-zero exit status. E: ./shoc: 3: ./bin/shocdriver: not found

Target: OpenCL - Benchmark: Bus Speed Download

dnn: The test quit with a non-zero exit status. E: ./shoc: 3: ./bin/shocdriver: not found

Target: OpenCL - Benchmark: Bus Speed Readback

dnn: The test quit with a non-zero exit status. E: ./shoc: 3: ./bin/shocdriver: not found

Target: OpenCL - Benchmark: Texture Read Bandwidth

dnn: The test quit with a non-zero exit status. E: ./shoc: 3: ./bin/shocdriver: not found

Libplacebo

Libplacebo is a multimedia rendering library based on the core rendering code of the MPV player. The libplacebo benchmark relies on the Vulkan API and tests various primitives. Learn more via the OpenBenchmarking.org test page.

dnn: The test quit with a non-zero exit status. E: fatal: Failed initializing vulkan device

cl-mem

A basic OpenCL memory benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: Copydnn90180270360450SE +/- 4.80, N = 15392.91. (CC) gcc options: -O2 -flto -lOpenCL

OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: Readdnn2004006008001000SE +/- 1.47, N = 3886.11. (CC) gcc options: -O2 -flto -lOpenCL

OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: Writednn2004006008001000SE +/- 1.50, N = 3801.61. (CC) gcc options: -O2 -flto -lOpenCL

NAMD CUDA

NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. This version of the NAMD test profile uses CUDA GPU acceleration. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD CUDA 2.14ATPase Simulation - 327,506 Atomsdnn0.01660.03320.04980.06640.083SE +/- 0.00099, N = 150.07371

Betsy GPU Compressor

Betsy is an open-source GPU compressor of various GPU compression techniques. Betsy is written in GLSL for Vulkan/OpenGL (compute shader) support for GPU-based texture compression. Learn more via the OpenBenchmarking.org test page.

Codec: ETC1 - Quality: Highest

dnn: The test quit with a non-zero exit status. E: ./betsy: 3: ./betsy: not found

Codec: ETC2 RGB - Quality: Highest

dnn: The test quit with a non-zero exit status. E: ./betsy: 3: ./betsy: not found

VkResample

VkResample is a Vulkan-based image upscaling library based on VkFFT. The sample input file is upscaling a 4K image to 8K using Vulkan-based GPU acceleration. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterVkResample 1.0Upscale: 2x - Precision: Doublednn1224364860SE +/- 0.03, N = 355.281. (CXX) g++ options: -O3

OpenBenchmarking.orgms, Fewer Is BetterVkResample 1.0Upscale: 2x - Precision: Singlednn246810SE +/- 0.008, N = 37.9111. (CXX) g++ options: -O3

OctaneBench

OctaneBench is a test of the OctaneRender on the GPU and requires the use of NVIDIA CUDA. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterOctaneBench 2020.1Total Scorednn300600900120015001281.89

RedShift Demo

This is a test of MAXON's RedShift demo build that currently requires NVIDIA GPU acceleration. Learn more via the OpenBenchmarking.org test page.

dnn: The test quit with a non-zero exit status. E: ./redshift: 3: /usr/redshift/bin/redshiftBenchmark: not found

FAHBench

FAHBench is a Folding@Home benchmark on the GPU. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNs Per Day, More Is BetterFAHBench 2.3.2dnn90180270360450SE +/- 0.68, N = 3430.65

clpeak

Clpeak is designed to test the peak capabilities of OpenCL devices. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGIOPS, More Is Betterclpeak 1.1.2OpenCL Test: Integer Compute INTdnn9K18K27K36K45KSE +/- 461.36, N = 340347.551. (CXX) g++ options: -O3

OpenBenchmarking.orgGFLOPS, More Is Betterclpeak 1.1.2OpenCL Test: Single-Precision Floatdnn20K40K60K80K100KSE +/- 2.35, N = 379707.101. (CXX) g++ options: -O3

OpenBenchmarking.orgGFLOPS, More Is Betterclpeak 1.1.2OpenCL Test: Double-Precision Doublednn30060090012001500SE +/- 2.09, N = 31396.961. (CXX) g++ options: -O3

OpenBenchmarking.orgGBPS, More Is Betterclpeak 1.1.2OpenCL Test: Global Memory Bandwidthdnn2004006008001000SE +/- 0.04, N = 3873.141. (CXX) g++ options: -O3

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: OpenCLdnn7K14K21K28K35KSE +/- 315.85, N = 3311991. (CXX) g++ options: -flto -pthread

Rodinia

Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes select OpenCL, NVIDIA CUDA and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 3.1Test: OpenCL Particle Filterdnn0.48890.97781.46671.95562.44452.1731. (CXX) g++ options: -O2 -lOpenCL

ArrayFire

ArrayFire is an GPU and CPU numeric processing library, this test uses the built-in CPU and OpenCL ArrayFire benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterArrayFire 3.7Test: Conjugate Gradient OpenCLdnn0.20890.41780.62670.83561.0445SE +/- 0.0086, N = 70.92861. (CXX) g++ options: -rdynamic

LuxCoreRender

LuxCoreRender is an open-source 3D physically based renderer formerly known as LuxRender. LuxCoreRender supports CPU-based rendering as well as GPU acceleration via OpenCL, NVIDIA CUDA, and NVIDIA OptiX interfaces. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: DLSC - Acceleration: GPUdnn612182430SE +/- 0.01, N = 325.83MIN: 24.42 / MAX: 26.13

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Danish Mood - Acceleration: GPUdnn510152025SE +/- 0.12, N = 319.98MIN: 7.42 / MAX: 23.23

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Orange Juice - Acceleration: GPUdnn510152025SE +/- 0.08, N = 320.19MIN: 18.16 / MAX: 27.9

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: LuxCore Benchmark - Acceleration: GPUdnn510152025SE +/- 0.04, N = 320.80MIN: 9.16 / MAX: 25.17

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.6Scene: Rainbow Colors and Prism - Acceleration: GPUdnn1020304050SE +/- 0.03, N = 344.93MIN: 38.13 / MAX: 47.42

FinanceBench

FinanceBench is a collection of financial program benchmarks with support for benchmarking on the GPU via OpenCL and CPU benchmarking with OpenMP. The FinanceBench test cases are focused on Black-Sholes-Merton Process with Analytic European Option engine, QMC (Sobol) Monte-Carlo method (Equity Option Example), Bonds Fixed-rate bond with flat forward curve, and Repo Securities repurchase agreement. FinanceBench was originally written by the Cavazos Lab at University of Delaware. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterFinanceBench 2016-07-25Benchmark: Black-Scholes OpenCLdnn0.66761.33522.00282.67043.338SE +/- 0.031, N = 32.9671. (CXX) g++ options: -O3 -march=native -fopenmp

ViennaCL

ViennaCL is an open-source linear algebra library written in C++ and with support for OpenCL and OpenMP. This test profile makes use of ViennaCL's built-in benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - sCOPYdnn4080120160200SE +/- 1.86, N = 32051. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - sAXPYdnn70140210280350SE +/- 3.51, N = 33091. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - sDOTdnn60120180240300SE +/- 4.26, N = 32941. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dCOPYdnn1428425670SE +/- 0.79, N = 362.41. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dAXPYdnn20406080100SE +/- 0.80, N = 392.91. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dDOTdnn20406080100SE +/- 0.72, N = 395.21. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMV-Ndnn20406080100SE +/- 0.33, N = 31111. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMV-Tdnn306090120150SE +/- 1.33, N = 31271. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMM-NNdnn1428425670SE +/- 0.51, N = 363.21. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMM-NTdnn1428425670SE +/- 0.59, N = 360.71. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMM-TNdnn1632486480SE +/- 1.29, N = 370.91. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: CPU BLAS - dGEMM-TTdnn1530456075SE +/- 0.44, N = 366.41. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - sCOPYdnn90180270360450SE +/- 0.88, N = 34361. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - sAXPYdnn120240360480600SE +/- 0.58, N = 35681. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dCOPYdnn140280420560700SE +/- 0.58, N = 36511. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dAXPYdnn160320480640800SE +/- 1.00, N = 37631. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dDOTdnn140280420560700SE +/- 6.74, N = 36421. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dGEMV-Ndnn50100150200250SE +/- 0.33, N = 32181. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dGEMV-Tdnn90180270360450SE +/- 0.88, N = 34351. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dGEMM-NNdnn2004006008001000SE +/- 0.00, N = 311601. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dGEMM-NTdnn30060090012001500SE +/- 3.33, N = 312771. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dGEMM-TNdnn30060090012001500SE +/- 3.33, N = 312971. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dGEMM-TTdnn30060090012001500SE +/- 3.33, N = 313471. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - sDOTdnn100200300400500SE +/- 0.50, N = 24391. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

GROMACS

The GROMACS (GROningen MAchine for Chemical Simulations) molecular dynamics package testing with the water_GMX50 data. This test profile allows selecting between CPU and GPU-based GROMACS builds. Learn more via the OpenBenchmarking.org test page.

Implementation: NVIDIA CUDA GPU - Input: water_GMX50_bare

dnn: The test quit with a non-zero exit status. E: /cuda-build/run-gromacs: 3: /cuda-build/bin/gmx: not found

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.

Model: AlexNet - Acceleration: NVIDIA CUDA - Iterations: 100

dnn: The test quit with a non-zero exit status. E: ./caffe: 3: ./tools/caffe: not found

Model: AlexNet - Acceleration: NVIDIA CUDA - Iterations: 200

dnn: The test quit with a non-zero exit status. E: ./caffe: 3: ./tools/caffe: not found

Model: AlexNet - Acceleration: NVIDIA CUDA - Iterations: 1000

dnn: The test quit with a non-zero exit status. E: ./caffe: 3: ./tools/caffe: not found

Model: GoogleNet - Acceleration: NVIDIA CUDA - Iterations: 100

dnn: The test quit with a non-zero exit status. E: ./caffe: 3: ./tools/caffe: not found

Model: GoogleNet - Acceleration: NVIDIA CUDA - Iterations: 200

dnn: The test quit with a non-zero exit status. E: ./caffe: 3: ./tools/caffe: not found

Model: GoogleNet - Acceleration: NVIDIA CUDA - Iterations: 1000

dnn: The test quit with a non-zero exit status. E: ./caffe: 3: ./tools/caffe: not found

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: Vulkan GPU - Model: mobilenetdnn0.68181.36362.04542.72723.409SE +/- 0.02, N = 33.03MIN: 2.23 / MAX: 70.111. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2dnn0.22950.4590.68850.9181.1475SE +/- 0.10, N = 31.02MIN: 0.76 / MAX: 3.431. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3dnn0.29030.58060.87091.16121.4515SE +/- 0.03, N = 31.29MIN: 1.12 / MAX: 7.861. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: shufflenet-v2dnn0.27680.55360.83041.10721.384SE +/- 0.02, N = 31.23MIN: 1.11 / MAX: 2.431. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: mnasnetdnn0.22730.45460.68190.90921.1365SE +/- 0.03, N = 31.01MIN: 0.87 / MAX: 4.41. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: efficientnet-b0dnn0.43650.8731.30951.7462.1825SE +/- 0.11, N = 31.94MIN: 1.56 / MAX: 72.481. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: blazefacednn0.1710.3420.5130.6840.855SE +/- 0.06, N = 30.76MIN: 0.6 / MAX: 2.671. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: googlenetdnn0.37580.75161.12741.50321.879SE +/- 0.08, N = 31.67MIN: 1.27 / MAX: 62.381. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: vgg16dnn0.3780.7561.1341.5121.89SE +/- 0.12, N = 21.68MIN: 1.45 / MAX: 67.031. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: alexnetdnn0.22950.4590.68850.9181.1475SE +/- 0.05, N = 31.02MIN: 0.89 / MAX: 4.191. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: yolov4-tinydnn1.01032.02063.03094.04125.0515SE +/- 0.08, N = 34.49MIN: 3.65 / MAX: 66.581. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: squeezenet_ssddnn0.71331.42662.13992.85323.5665SE +/- 0.22, N = 33.17MIN: 2.32 / MAX: 68.21. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: regnety_400mdnn0.33750.6751.01251.351.6875SE +/- 0.06, N = 31.50MIN: 1.31 / MAX: 3.731. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: vision_transformerdnn306090120150SE +/- 7.90, N = 3121.14MIN: 89.4 / MAX: 315.411. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: FastestDetdnn0.33080.66160.99241.32321.654SE +/- 0.12, N = 31.47MIN: 1.28 / MAX: 64.751. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: resnet18dnn0.19350.3870.58050.7740.9675SE +/- 0.06, N = 20.86MIN: 0.76 / MAX: 3.951. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: resnet50dnn0.3420.6841.0261.3681.711.52MIN: 1.43 / MAX: 2.441. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

PlaidML

This test profile uses PlaidML deep learning framework developed by Intel for offering up various benchmarks. Learn more via the OpenBenchmarking.org test page.

FP16: No - Mode: Training - Network: Mobilenet - Device: OpenCL

dnn: The test quit with a non-zero exit status. E: ImportError: cannot import name 'Iterable' from 'collections' (/usr/lib/python3.10/collections/__init__.py)

FP16: No - Mode: Inference - Network: IMDB LSTM - Device: OpenCL

dnn: The test quit with a non-zero exit status. E: ImportError: cannot import name 'Iterable' from 'collections' (/usr/lib/python3.10/collections/__init__.py)

FP16: No - Mode: Inference - Network: Mobilenet - Device: OpenCL

dnn: The test quit with a non-zero exit status. E: ImportError: cannot import name 'Iterable' from 'collections' (/usr/lib/python3.10/collections/__init__.py)

FP16: Yes - Mode: Inference - Network: Mobilenet - Device: OpenCL

dnn: The test quit with a non-zero exit status. E: ImportError: cannot import name 'Iterable' from 'collections' (/usr/lib/python3.10/collections/__init__.py)

FP16: No - Mode: Inference - Network: DenseNet 201 - Device: OpenCL

dnn: The test quit with a non-zero exit status. E: ImportError: cannot import name 'Iterable' from 'collections' (/usr/lib/python3.10/collections/__init__.py)

Blender

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

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.5Blend File: BMW27 - Compute: NVIDIA OptiXdnn3691215SE +/- 9.91, N = 1213.47

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.5Blend File: Classroom - Compute: NVIDIA OptiXdnn246810SE +/- 0.03, N = 37.34

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.5Blend File: Fishy Cat - Compute: NVIDIA OptiXdnn1.25552.5113.76655.0226.2775SE +/- 0.05, N = 145.58

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.5Blend File: Barbershop - Compute: NVIDIA OptiXdnn714212835SE +/- 0.08, N = 330.60

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.5Blend File: Pabellon Barcelona - Compute: NVIDIA OptiXdnn246810SE +/- 0.03, N = 38.37

IndigoBench

This is a test of Indigo Renderer's IndigoBench benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/s, More Is BetterIndigoBench 4.4Acceleration: OpenCL GPU - Scene: Bedroomdnn816243240SE +/- 0.02, N = 335.51

OpenBenchmarking.orgM samples/s, More Is BetterIndigoBench 4.4Acceleration: OpenCL GPU - Scene: Supercardnn20406080100SE +/- 0.01, N = 379.45

MandelGPU

MandelGPU is an OpenCL benchmark and this test runs with the OpenCL rendering float4 kernel with a maximum of 4096 iterations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSamples/sec, More Is BetterMandelGPU 1.3pts1OpenCL Device: GPUdnn200M400M600M800M1000MSE +/- 1842939.53, N = 3830513186.51. (CC) gcc options: -O3 -lm -ftree-vectorize -funroll-loops -lglut -lOpenCL -lGL

NeatBench

NeatBench is a benchmark of the cross-platform Neat Video software on the CPU and optional GPU (OpenCL / CUDA) support. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterNeatBench 5Acceleration: GPUdnn9001800270036004500SE +/- 0.00, N = 34090

85 Results Shown

vkpeak:
  fp32-scalar
  fp32-vec4
  fp16-scalar
  fp16-vec4
  fp64-scalar
  fp64-vec4
  int32-scalar
  int32-vec4
  int16-scalar
  int16-vec4
RealSR-NCNN:
  4x - No
  4x - Yes
Waifu2x-NCNN Vulkan
VkFFT
cl-mem:
  Copy
  Read
  Write
NAMD CUDA
VkResample:
  2x - Double
  2x - Single
OctaneBench
FAHBench
clpeak:
  Integer Compute INT
  Single-Precision Float
  Double-Precision Double
  Global Memory Bandwidth
LeelaChessZero
Rodinia
ArrayFire
LuxCoreRender:
  DLSC - GPU
  Danish Mood - GPU
  Orange Juice - GPU
  LuxCore Benchmark - GPU
  Rainbow Colors and Prism - GPU
FinanceBench
ViennaCL:
  CPU BLAS - sCOPY
  CPU BLAS - sAXPY
  CPU BLAS - sDOT
  CPU BLAS - dCOPY
  CPU BLAS - dAXPY
  CPU BLAS - dDOT
  CPU BLAS - dGEMV-N
  CPU BLAS - dGEMV-T
  CPU BLAS - dGEMM-NN
  CPU BLAS - dGEMM-NT
  CPU BLAS - dGEMM-TN
  CPU BLAS - dGEMM-TT
  OpenCL BLAS - sCOPY
  OpenCL BLAS - sAXPY
  OpenCL BLAS - dCOPY
  OpenCL BLAS - dAXPY
  OpenCL BLAS - dDOT
  OpenCL BLAS - dGEMV-N
  OpenCL BLAS - dGEMV-T
  OpenCL BLAS - dGEMM-NN
  OpenCL BLAS - dGEMM-NT
  OpenCL BLAS - dGEMM-TN
  OpenCL BLAS - dGEMM-TT
  OpenCL BLAS - sDOT
NCNN:
  Vulkan GPU - mobilenet
  Vulkan GPU-v2-v2 - mobilenet-v2
  Vulkan GPU-v3-v3 - mobilenet-v3
  Vulkan GPU - shufflenet-v2
  Vulkan GPU - mnasnet
  Vulkan GPU - efficientnet-b0
  Vulkan GPU - blazeface
  Vulkan GPU - googlenet
  Vulkan GPU - vgg16
  Vulkan GPU - alexnet
  Vulkan GPU - yolov4-tiny
  Vulkan GPU - squeezenet_ssd
  Vulkan GPU - regnety_400m
  Vulkan GPU - vision_transformer
  Vulkan GPU - FastestDet
  Vulkan GPU - resnet18
  Vulkan GPU - resnet50
Blender:
  BMW27 - NVIDIA OptiX
  Classroom - NVIDIA OptiX
  Fishy Cat - NVIDIA OptiX
  Barbershop - NVIDIA OptiX
  Pabellon Barcelona - NVIDIA OptiX
IndigoBench:
  OpenCL GPU - Bedroom
  OpenCL GPU - Supercar
MandelGPU
NeatBench