NVIDIA GPU Compute Benchmarks

Benchmarks 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 2106127-IB-3080C630035
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
Performance Per
Dollar
Date
Run
  Test
  Duration
RTX 3080 RBAR
June 12 2021
  3 Hours, 54 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):

RTX 3080 RBAR

Processor: AMD Ryzen 9 5900X 12-Core @ 3.70GHz (12 Cores / 24 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (3402 BIOS), Chipset: AMD Starship/Matisse, Memory: 16GB, Disk: 1000GB Sabrent Rocket 4.0 Plus + 2000GB, Graphics: NVIDIA GeForce RTX 3080 10GB, Audio: NVIDIA GA102 HD Audio, Monitor: ASUS VP28U, Network: Realtek RTL8125 2.5GbE + Intel I211

OS: Ubuntu 21.04, Kernel: 5.11.0-17-generic (x86_64), Desktop: GNOME Shell 3.38.4, Display Server: X Server 1.20.11, Display Driver: NVIDIA 465.31, OpenGL: 4.6.0, OpenCL: OpenCL 3.0 CUDA 11.3.116, Vulkan: 1.2.168, Compiler: GCC 10.3.0 + CUDA 11.3, File-System: ext4, Screen Resolution: 3840x2160

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --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-mutex --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Processor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa201009
OpenCL Notes: GPU Compute Cores: 8704
Python Notes: Python 3.9.5
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected

NVIDIA GPU Compute BenchmarksOpenBenchmarking.orgPhoronix Test SuiteAMD Ryzen 9 5900X 12-Core @ 3.70GHz (12 Cores / 24 Threads)ASUS ROG CROSSHAIR VIII HERO (3402 BIOS)AMD Starship/Matisse16GB1000GB Sabrent Rocket 4.0 Plus + 2000GBNVIDIA GeForce RTX 3080 10GBNVIDIA GA102 HD AudioASUS VP28URealtek RTL8125 2.5GbE + Intel I211Ubuntu 21.045.11.0-17-generic (x86_64)GNOME Shell 3.38.4X Server 1.20.11NVIDIA 465.314.6.0OpenCL 3.0 CUDA 11.3.1161.2.168GCC 10.3.0 + CUDA 11.3ext43840x2160ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLOpenCLVulkanCompilerFile-SystemScreen ResolutionNVIDIA GPU Compute Benchmarks PerformanceSystem Logs- Transparent Huge Pages: madvise- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --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-mutex --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v - Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa201009 - GPU Compute Cores: 8704- Python 3.9.5- itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected

NVIDIA GPU Compute Benchmarksviennacl: OpenCL BLAS - dGEMM-TTviennacl: OpenCL BLAS - dGEMV-Tviennacl: OpenCL BLAS - dGEMV-Nviennacl: OpenCL BLAS - sDOTredshift: betsy: ETC2 RGB - Highestbetsy: ETC1 - Highestwaifu2x-ncnn: 2x - 3 - Yesrealsr-ncnn: 4x - Norealsr-ncnn: 4x - Yesvkresample: 2x - Singlev-ray: NVIDIA RTX GPUv-ray: NVIDIA CUDA GPUblender: Barbershop - NVIDIA OptiXblender: Barbershop - CUDAblender: Pabellon Barcelona - NVIDIA OptiXblender: Pabellon Barcelona - CUDAblender: Fishy Cat - NVIDIA OptiXblender: Fishy Cat - CUDAblender: Classroom - NVIDIA OptiXblender: Classroom - CUDAblender: BMW27 - NVIDIA OptiXblender: BMW27 - CUDAindigobench: OpenCL GPU - Bedroomindigobench: OpenCL GPU - Supercarshoc: OpenCL - S3Dshoc: OpenCL - MD5 Hashshoc: OpenCL - GEMM SGEMM_Nshoc: OpenCL - FFT SPshoc: OpenCL - Texture Read Bandwidthviennacl: OpenCL BLAS - dGEMM-TNviennacl: OpenCL BLAS - dGEMM-NTviennacl: OpenCL BLAS - dGEMM-NNviennacl: OpenCL BLAS - dDOTviennacl: OpenCL BLAS - dAXPYviennacl: OpenCL BLAS - dCOPYviennacl: OpenCL BLAS - sAXPYviennacl: OpenCL BLAS - sCOPYcl-mem: Copycl-mem: Writecl-mem: Readlczero: OpenCLplaidml: No - Inference - VGG19 - OpenCLplaidml: No - Inference - VGG16 - OpenCLplaidml: No - Inference - ResNet 50 - OpenCLvkpeak: int16-vec4vkpeak: int16-scalarvkpeak: int32-vec4vkpeak: int32-scalarvkpeak: fp64-vec4vkpeak: fp64-scalarvkpeak: fp16-vec4vkpeak: fp16-scalarvkpeak: fp32-vec4vkpeak: fp32-scalarclpeak: Integer Compute INTclpeak: Double-Precision Doubleclpeak: Single-Precision Floatclpeak: Global Memory Bandwidtharrayfire: Conjugate Gradient OpenCLarrayfire: BLAS OpenCLluxcorerender: Danish Mood - GPUluxcorerender: Orange Juice - GPUluxcorerender: LuxCore Benchmark - GPUluxcorerender: Rainbow Colors and Prism - GPUluxcorerender: DLSC - GPUoctanebench: Total Scorenamd-cuda: ATPase Simulation - 327,506 Atomsfahbench: hashcat: TrueCrypt RIPEMD160 + XTShashcat: 7-Ziphashcat: SHA-512hashcat: SHA1hashcat: MD5mixbench: NVIDIA CUDA - Integermixbench: NVIDIA CUDA - Half Precisionmixbench: NVIDIA CUDA - Double Precisionmixbench: NVIDIA CUDA - Single PrecisionRTX 3080 RBAR4963742403631654.5283.3443.3866.11532.57011.30423791773408.75445.3953.33147.0222.9841.6936.4161.2211.3622.1117.47246.144337.64337.21726243.801914.112205.99494498496596627552477354353.4634.5670.337359223.01280.78727.4114506.5211282.6717007.0317069.59536.05534.1034257.5517075.8022762.9617141.9215383.91540.4429422.68659.401.5637394.208.0310.879.9828.4911.62567.0265560.12157321.05467201679957332753700000191645333335859836666713924.2031706.66417.9829095.86OpenBenchmarking.org

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.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dGEMM-TTRTX 3080 RBAR1102203304405504961. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dGEMV-TRTX 3080 RBAR80160240320400SE +/- 0.33, N = 33741. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dGEMV-NRTX 3080 RBAR50100150200250SE +/- 0.33, N = 32401. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - sDOTRTX 3080 RBAR80160240320400SE +/- 0.33, N = 33631. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

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.

OpenBenchmarking.orgSeconds, Fewer Is BetterRedShift Demo 3.0RTX 3080 RBAR4080120160200SE +/- 0.67, N = 3165

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.

OpenBenchmarking.orgSeconds, Fewer Is BetterBetsy GPU Compressor 1.1 BetaCodec: ETC2 RGB - Quality: HighestRTX 3080 RBAR1.01882.03763.05644.07525.094SE +/- 0.049, N = 34.5281. (CXX) g++ options: -O3 -O2 -lpthread -ldl

OpenBenchmarking.orgSeconds, Fewer Is BetterBetsy GPU Compressor 1.1 BetaCodec: ETC1 - Quality: HighestRTX 3080 RBAR0.75241.50482.25723.00963.762SE +/- 0.015, N = 33.3441. (CXX) g++ options: -O3 -O2 -lpthread -ldl

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.

OpenBenchmarking.orgSeconds, Fewer Is BetterWaifu2x-NCNN Vulkan 20200818Scale: 2x - Denoise: 3 - TAA: YesRTX 3080 RBAR0.76191.52382.28573.04763.8095SE +/- 0.006, N = 33.386

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: NoRTX 3080 RBAR246810SE +/- 0.009, N = 36.115

OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: YesRTX 3080 RBAR816243240SE +/- 0.07, N = 332.57

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: SingleRTX 3080 RBAR3691215SE +/- 0.01, N = 311.301. (CXX) g++ options: -O3 -pthread

Chaos Group V-RAY

This is a test of Chaos Group's V-RAY benchmark. V-RAY is a commercial renderer that can integrate with various creator software products like SketchUp and 3ds Max. The V-RAY benchmark is standalone and supports CPU and NVIDIA CUDA/RTX based rendering. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgvrays, More Is BetterChaos Group V-RAY 5Mode: NVIDIA RTX GPURTX 3080 RBAR5001000150020002500SE +/- 2.73, N = 32379

OpenBenchmarking.orgvpaths, More Is BetterChaos Group V-RAY 5Mode: NVIDIA CUDA GPURTX 3080 RBAR400800120016002000SE +/- 0.33, N = 31773

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL, NVIDIA OptiX, and NVIDIA CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Barbershop - Compute: NVIDIA OptiXRTX 3080 RBAR90180270360450SE +/- 0.62, N = 3408.75

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Barbershop - Compute: CUDARTX 3080 RBAR100200300400500SE +/- 0.22, N = 3445.39

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Pabellon Barcelona - Compute: NVIDIA OptiXRTX 3080 RBAR1224364860SE +/- 0.17, N = 353.33

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Pabellon Barcelona - Compute: CUDARTX 3080 RBAR306090120150SE +/- 0.06, N = 3147.02

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Fishy Cat - Compute: NVIDIA OptiXRTX 3080 RBAR612182430SE +/- 0.02, N = 322.98

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Fishy Cat - Compute: CUDARTX 3080 RBAR1020304050SE +/- 0.01, N = 341.69

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Classroom - Compute: NVIDIA OptiXRTX 3080 RBAR816243240SE +/- 0.03, N = 336.41

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Classroom - Compute: CUDARTX 3080 RBAR1428425670SE +/- 0.08, N = 361.22

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: BMW27 - Compute: NVIDIA OptiXRTX 3080 RBAR3691215SE +/- 0.03, N = 311.36

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: BMW27 - Compute: CUDARTX 3080 RBAR510152025SE +/- 0.03, N = 322.11

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: BedroomRTX 3080 RBAR48121620SE +/- 0.01, N = 317.47

OpenBenchmarking.orgM samples/s, More Is BetterIndigoBench 4.4Acceleration: OpenCL GPU - Scene: SupercarRTX 3080 RBAR1020304050SE +/- 0.04, N = 346.14

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.

OpenBenchmarking.orgGFLOPS, More Is BetterSHOC Scalable HeterOgeneous Computing 2020-04-17Target: OpenCL - Benchmark: S3DRTX 3080 RBAR70140210280350SE +/- 0.27, N = 3337.641. (CXX) g++ options: -O2 -lSHOCCommonMPI -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt -pthread -lmpi_cxx -lmpi

OpenBenchmarking.orgGHash/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2020-04-17Target: OpenCL - Benchmark: MD5 HashRTX 3080 RBAR918273645SE +/- 0.00, N = 337.221. (CXX) g++ options: -O2 -lSHOCCommonMPI -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt -pthread -lmpi_cxx -lmpi

OpenBenchmarking.orgGFLOPS, More Is BetterSHOC Scalable HeterOgeneous Computing 2020-04-17Target: OpenCL - Benchmark: GEMM SGEMM_NRTX 3080 RBAR13002600390052006500SE +/- 27.64, N = 36243.801. (CXX) g++ options: -O2 -lSHOCCommonMPI -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt -pthread -lmpi_cxx -lmpi

OpenBenchmarking.orgGFLOPS, More Is BetterSHOC Scalable HeterOgeneous Computing 2020-04-17Target: OpenCL - Benchmark: FFT SPRTX 3080 RBAR400800120016002000SE +/- 3.10, N = 31914.111. (CXX) g++ options: -O2 -lSHOCCommonMPI -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt -pthread -lmpi_cxx -lmpi

OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2020-04-17Target: OpenCL - Benchmark: Texture Read BandwidthRTX 3080 RBAR5001000150020002500SE +/- 3.47, N = 32205.991. (CXX) g++ options: -O2 -lSHOCCommonMPI -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt -pthread -lmpi_cxx -lmpi

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.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dGEMM-TNRTX 3080 RBAR110220330440550SE +/- 1.00, N = 34941. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dGEMM-NTRTX 3080 RBAR110220330440550SE +/- 1.20, N = 34981. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPs/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dGEMM-NNRTX 3080 RBAR110220330440550SE +/- 1.33, N = 34961. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dDOTRTX 3080 RBAR1302603905206505961. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dAXPYRTX 3080 RBAR1402804205607006271. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dCOPYRTX 3080 RBAR120240360480600SE +/- 0.33, N = 35521. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - sAXPYRTX 3080 RBAR100200300400500SE +/- 0.33, N = 34771. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - sCOPYRTX 3080 RBAR80160240320400SE +/- 0.67, N = 33541. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

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: CopyRTX 3080 RBAR80160240320400SE +/- 0.12, N = 3353.41. (CC) gcc options: -O2 -flto -lOpenCL

OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: WriteRTX 3080 RBAR140280420560700SE +/- 0.68, N = 3634.51. (CC) gcc options: -O2 -flto -lOpenCL

OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: ReadRTX 3080 RBAR140280420560700SE +/- 0.82, N = 3670.31. (CC) gcc options: -O2 -flto -lOpenCL

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 0.26Backend: OpenCLRTX 3080 RBAR8K16K24K32K40KSE +/- 129.19, N = 3373591. (CXX) g++ options: -flto -pthread

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.

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: VGG19 - Device: OpenCLRTX 3080 RBAR50100150200250SE +/- 0.43, N = 3223.01

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: VGG16 - Device: OpenCLRTX 3080 RBAR60120180240300SE +/- 0.65, N = 3280.78

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: ResNet 50 - Device: OpenCLRTX 3080 RBAR160320480640800SE +/- 1.15, N = 3727.41

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.orgGIOPS, More Is Bettervkpeak 20210424int16-vec4RTX 3080 RBAR3K6K9K12K15KSE +/- 19.69, N = 314506.52

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int16-scalarRTX 3080 RBAR2K4K6K8K10KSE +/- 0.12, N = 311282.67

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int32-vec4RTX 3080 RBAR4K8K12K16K20KSE +/- 2.37, N = 317007.03

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int32-scalarRTX 3080 RBAR4K8K12K16K20KSE +/- 3.48, N = 317069.59

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp64-vec4RTX 3080 RBAR120240360480600SE +/- 0.35, N = 3536.05

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp64-scalarRTX 3080 RBAR120240360480600SE +/- 1.41, N = 3534.10

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp16-vec4RTX 3080 RBAR7K14K21K28K35KSE +/- 84.67, N = 334257.55

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp16-scalarRTX 3080 RBAR4K8K12K16K20KSE +/- 42.62, N = 317075.80

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp32-vec4RTX 3080 RBAR5K10K15K20K25KSE +/- 39.23, N = 322762.96

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp32-scalarRTX 3080 RBAR4K8K12K16K20KSE +/- 42.89, N = 317141.92

clpeak

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

OpenBenchmarking.orgGIOPS, More Is BetterclpeakOpenCL Test: Integer Compute INTRTX 3080 RBAR3K6K9K12K15KSE +/- 105.69, N = 1515383.911. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Double-Precision DoubleRTX 3080 RBAR120240360480600SE +/- 0.39, N = 3540.441. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Single-Precision FloatRTX 3080 RBAR6K12K18K24K30KSE +/- 169.65, N = 329422.681. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Global Memory BandwidthRTX 3080 RBAR140280420560700SE +/- 2.74, N = 3659.401. (CXX) g++ options: -O3 -rdynamic -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 OpenCLRTX 3080 RBAR0.35170.70341.05511.40681.7585SE +/- 0.003, N = 31.5631. (CXX) g++ options: -rdynamic

OpenBenchmarking.orgGFLOPS, More Is BetterArrayFire 3.7Test: BLAS OpenCLRTX 3080 RBAR16003200480064008000SE +/- 3.09, N = 37394.201. (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 2.5Scene: Danish Mood - Acceleration: GPURTX 3080 RBAR246810SE +/- 0.06, N = 38.03MIN: 2.38 / MAX: 9.73

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Orange Juice - Acceleration: GPURTX 3080 RBAR3691215SE +/- 0.01, N = 310.87MIN: 8.73 / MAX: 14.44

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: LuxCore Benchmark - Acceleration: GPURTX 3080 RBAR3691215SE +/- 0.01, N = 39.98MIN: 3.02 / MAX: 11.81

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Rainbow Colors and Prism - Acceleration: GPURTX 3080 RBAR714212835SE +/- 0.06, N = 328.49MIN: 26.83 / MAX: 31.13

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: DLSC - Acceleration: GPURTX 3080 RBAR3691215SE +/- 0.01, N = 311.62MIN: 11.21 / MAX: 11.85

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 ScoreRTX 3080 RBAR120240360480600567.03

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 AtomsRTX 3080 RBAR0.02740.05480.08220.10960.137SE +/- 0.00003, N = 30.12157

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 2.3.2RTX 3080 RBAR70140210280350SE +/- 0.67, N = 3321.05

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.

OpenBenchmarking.orgH/s, More Is BetterHashcat 6.1.1Benchmark: TrueCrypt RIPEMD160 + XTSRTX 3080 RBAR150K300K450K600K750KSE +/- 1322.04, N = 3720167

OpenBenchmarking.orgH/s, More Is BetterHashcat 6.1.1Benchmark: 7-ZipRTX 3080 RBAR200K400K600K800K1000KSE +/- 2771.48, N = 3995733

OpenBenchmarking.orgH/s, More Is BetterHashcat 6.1.1Benchmark: SHA-512RTX 3080 RBAR600M1200M1800M2400M3000MSE +/- 2042057.79, N = 32753700000

OpenBenchmarking.orgH/s, More Is BetterHashcat 6.1.1Benchmark: SHA1RTX 3080 RBAR4000M8000M12000M16000M20000MSE +/- 22857116.57, N = 319164533333

OpenBenchmarking.orgH/s, More Is BetterHashcat 6.1.1Benchmark: MD5RTX 3080 RBAR13000M26000M39000M52000M65000MSE +/- 81551462.96, N = 358598366667

Mixbench

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

OpenBenchmarking.orgGIOPS, More Is BetterMixbench 2020-06-23Backend: NVIDIA CUDA - Benchmark: IntegerRTX 3080 RBAR3K6K9K12K15KSE +/- 258.54, N = 1513924.201. (CXX) g++ options: -lm -lstdc++ -lOpenCL -lrt -O2

OpenBenchmarking.orgGFLOPS, More Is BetterMixbench 2020-06-23Backend: NVIDIA CUDA - Benchmark: Half PrecisionRTX 3080 RBAR7K14K21K28K35KSE +/- 650.08, N = 1231706.661. (CXX) g++ options: -lm -lstdc++ -lOpenCL -lrt -O2

OpenBenchmarking.orgGFLOPS, More Is BetterMixbench 2020-06-23Backend: NVIDIA CUDA - Benchmark: Double PrecisionRTX 3080 RBAR90180270360450SE +/- 7.75, N = 15417.981. (CXX) g++ options: -lm -lstdc++ -lOpenCL -lrt -O2

OpenBenchmarking.orgGFLOPS, More Is BetterMixbench 2020-06-23Backend: NVIDIA CUDA - Benchmark: Single PrecisionRTX 3080 RBAR6K12K18K24K30KSE +/- 555.26, N = 1529095.861. (CXX) g++ options: -lm -lstdc++ -lOpenCL -lrt -O2

78 Results Shown

ViennaCL:
  OpenCL BLAS - dGEMM-TT
  OpenCL BLAS - dGEMV-T
  OpenCL BLAS - dGEMV-N
  OpenCL BLAS - sDOT
RedShift Demo
Betsy GPU Compressor:
  ETC2 RGB - Highest
  ETC1 - Highest
Waifu2x-NCNN Vulkan
RealSR-NCNN:
  4x - No
  4x - Yes
VkResample
Chaos Group V-RAY:
  NVIDIA RTX GPU
  NVIDIA CUDA GPU
Blender:
  Barbershop - NVIDIA OptiX
  Barbershop - CUDA
  Pabellon Barcelona - NVIDIA OptiX
  Pabellon Barcelona - CUDA
  Fishy Cat - NVIDIA OptiX
  Fishy Cat - CUDA
  Classroom - NVIDIA OptiX
  Classroom - CUDA
  BMW27 - NVIDIA OptiX
  BMW27 - CUDA
IndigoBench:
  OpenCL GPU - Bedroom
  OpenCL GPU - Supercar
SHOC Scalable HeterOgeneous Computing:
  OpenCL - S3D
  OpenCL - MD5 Hash
  OpenCL - GEMM SGEMM_N
  OpenCL - FFT SP
  OpenCL - Texture Read Bandwidth
ViennaCL:
  OpenCL BLAS - dGEMM-TN
  OpenCL BLAS - dGEMM-NT
  OpenCL BLAS - dGEMM-NN
  OpenCL BLAS - dDOT
  OpenCL BLAS - dAXPY
  OpenCL BLAS - dCOPY
  OpenCL BLAS - sAXPY
  OpenCL BLAS - sCOPY
cl-mem:
  Copy
  Write
  Read
LeelaChessZero
PlaidML:
  No - Inference - VGG19 - OpenCL
  No - Inference - VGG16 - OpenCL
  No - Inference - ResNet 50 - OpenCL
vkpeak:
  int16-vec4
  int16-scalar
  int32-vec4
  int32-scalar
  fp64-vec4
  fp64-scalar
  fp16-vec4
  fp16-scalar
  fp32-vec4
  fp32-scalar
clpeak:
  Integer Compute INT
  Double-Precision Double
  Single-Precision Float
  Global Memory Bandwidth
ArrayFire:
  Conjugate Gradient OpenCL
  BLAS OpenCL
LuxCoreRender:
  Danish Mood - GPU
  Orange Juice - GPU
  LuxCore Benchmark - GPU
  Rainbow Colors and Prism - GPU
  DLSC - GPU
OctaneBench
NAMD CUDA
FAHBench
Hashcat:
  TrueCrypt RIPEMD160 + XTS
  7-Zip
  SHA-512
  SHA1
  MD5
Mixbench:
  NVIDIA CUDA - Integer
  NVIDIA CUDA - Half Precision
  NVIDIA CUDA - Double Precision
  NVIDIA CUDA - Single Precision

RTX 3080 RBAR

Processor: AMD Ryzen 9 5900X 12-Core @ 3.70GHz (12 Cores / 24 Threads), Motherboard: ASUS ROG CROSSHAIR VIII HERO (3402 BIOS), Chipset: AMD Starship/Matisse, Memory: 16GB, Disk: 1000GB Sabrent Rocket 4.0 Plus + 2000GB, Graphics: NVIDIA GeForce RTX 3080 10GB, Audio: NVIDIA GA102 HD Audio, Monitor: ASUS VP28U, Network: Realtek RTL8125 2.5GbE + Intel I211

OS: Ubuntu 21.04, Kernel: 5.11.0-17-generic (x86_64), Desktop: GNOME Shell 3.38.4, Display Server: X Server 1.20.11, Display Driver: NVIDIA 465.31, OpenGL: 4.6.0, OpenCL: OpenCL 3.0 CUDA 11.3.116, Vulkan: 1.2.168, Compiler: GCC 10.3.0 + CUDA 11.3, File-System: ext4, Screen Resolution: 3840x2160

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --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-mutex --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-gDeRY6/gcc-10-10.3.0/debian/tmp-gcn/usr,hsa --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Processor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa201009
OpenCL Notes: GPU Compute Cores: 8704
Python Notes: Python 3.9.5
Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 12 June 2021 11:11 by user phoronix.