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 2106089-IB-NVIDIACOM84
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:

BLAS (Basic Linear Algebra Sub-Routine) Tests 2 Tests
CPU Massive 4 Tests
Creator Workloads 5 Tests
Game Development 2 Tests
HPC - High Performance Computing 4 Tests
Machine Learning 3 Tests
Multi-Core 5 Tests
NVIDIA GPU Compute 22 Tests
OpenCL 4 Tests
Renderers 4 Tests
Vulkan Compute 5 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
Disable Color Branding
Prefer Vertical Bar Graphs
No Box Plots
On Line Graphs With Missing Data, Connect The Line Gaps

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs
Condense Test Profiles With Multiple Version Results 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
RTX 2060 SUPER
June 03 2021
  7 Hours, 7 Minutes
RTX 2070
June 06 2021
  7 Hours, 8 Minutes
RTX 2070 SUPER
June 05 2021
  6 Hours, 3 Minutes
RTX 2080
June 03 2021
  6 Hours, 9 Minutes
RTX 2080 SUPER
June 05 2021
  5 Hours, 28 Minutes
RTX 2080 Ti
June 06 2021
  4 Hours, 47 Minutes
TITAN RTX
June 04 2021
  4 Hours, 45 Minutes
RTX 3060
June 04 2021
  5 Hours, 58 Minutes
RTX 3060 Ti
June 05 2021
  5 Hours, 10 Minutes
RTX 3070
June 04 2021
  4 Hours, 37 Minutes
RTX 3070 Ti
June 02 2021
  4 Hours, 49 Minutes
RTX 3080
June 03 2021
  4 Hours, 1 Minute
RTX 3080 Ti
June 03 2021
  4 Hours, 36 Minutes
RTX 3090
June 02 2021
  4 Hours, 22 Minutes
Invert Hiding All Results Option
  5 Hours, 21 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):


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 2060 SUPER 8GBASUS NVIDIA GeForce RTX 2070 8GBNVIDIA GeForce RTX 2070 SUPER 8GBZotac NVIDIA GeForce RTX 2080 8GBNVIDIA GeForce RTX 2080 SUPER 8GBNVIDIA GeForce RTX 2080 Ti 11GBNVIDIA TITAN RTX 24GBeVGA NVIDIA GeForce RTX 3060 12GBNVIDIA GeForce RTX 3060 Ti 8GBNVIDIA GeForce RTX 3070 8GBNVIDIA GeForce RTX 3070 Ti 8GBNVIDIA GeForce RTX 3080 10GBNVIDIA GeForce RTX 3080 Ti 12GBNVIDIA GeForce RTX 3090 24GBNVIDIA TU106 HD AudioNVIDIA TU104 HD AudioNVIDIA TU102 HD AudioNVIDIA GeForce RTX 3060NVIDIA GA104 HD AudioNVIDIA 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 11.3ext43840x2160ProcessorMotherboardChipsetMemoryDiskGraphicsAudiosMonitorNetworkOSKernelDesktopDisplay 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 - RTX 2060 SUPER: GPU Compute Cores: 2176- RTX 2070: GPU Compute Cores: 2304- RTX 2070 SUPER: GPU Compute Cores: 2560- RTX 2080: GPU Compute Cores: 2944- RTX 2080 SUPER: GPU Compute Cores: 3072- RTX 2080 Ti: GPU Compute Cores: 4352- TITAN RTX: GPU Compute Cores: 4608- RTX 3060: GPU Compute Cores: 3584- RTX 3060 Ti: GPU Compute Cores: 4864- RTX 3070: GPU Compute Cores: 5888- RTX 3070 Ti: GPU Compute Cores: 6144- RTX 3080: GPU Compute Cores: 8704- RTX 3080 Ti: GPU Compute Cores: 10240- RTX 3090: GPU Compute Cores: 10496- 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

RTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiTITAN RTXRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090Result OverviewPhoronix Test Suite100%176%252%328%404%BlenderChaos Group V-RAYLuxCoreRendervkpeakclpeakHashcatOctaneBenchMixbenchLeelaChessZeroRedShift DemoPlaidMLIndigoBenchSHOC Scalable HeterOgeneous ComputingVkResampleRealSR-NCNNArrayFireViennaCLcl-memFAHBenchWaifu2x-NCNN VulkanNAMD CUDA

RTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiTITAN RTXRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090Per Watt Result OverviewPhoronix Test Suite100%121%141%162%Chaos Group V-RAYArrayFireMixbenchclpeakLuxCoreRenderLeelaChessZeroOctaneBenchIndigoBenchPlaidMLFAHBenchHashcatSHOC Scalable HeterOgeneous Computingcl-memP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.MP.W.G.M

NVIDIA GPU Compute Benchmarksblender: Classroom - CUDAblender: Barbershop - NVIDIA OptiXclpeak: Single-Precision Floatblender: Classroom - NVIDIA OptiXblender: Pabellon Barcelona - CUDAblender: Pabellon Barcelona - NVIDIA OptiXv-ray: NVIDIA CUDA GPUv-ray: NVIDIA RTX GPUluxcorerender: DLSC - GPUblender: Fishy Cat - CUDAblender: BMW27 - CUDAluxcorerender: Danish Mood - GPUvkpeak: fp32-vec4arrayfire: BLAS OpenCLblender: BMW27 - NVIDIA OptiXblender: Fishy Cat - NVIDIA OptiXluxcorerender: LuxCore Benchmark - GPUvkpeak: fp64-vec4vkpeak: fp32-scalarvkpeak: fp64-scalarclpeak: Double-Precision Doublevkpeak: int32-scalarvkpeak: int16-scalarvkpeak: fp16-scalarvkpeak: int32-vec4vkpeak: fp16-vec4shoc: OpenCL - GEMM SGEMM_Nblender: Barbershop - CUDAviennacl: OpenCL BLAS - dGEMM-NNhashcat: SHA1viennacl: OpenCL BLAS - dGEMM-TNhashcat: MD5shoc: OpenCL - MD5 Hashhashcat: SHA-512clpeak: Integer Compute INTvkpeak: int16-vec4hashcat: TrueCrypt RIPEMD160 + XTSindigobench: OpenCL GPU - Bedroomluxcorerender: Rainbow Colors and Prism - GPUmixbench: NVIDIA CUDA - Half Precisionoctanebench: Total Scoreluxcorerender: Orange Juice - GPUlczero: OpenCLredshift: realsr-ncnn: 4x - Yeshashcat: 7-Zipplaidml: No - Inference - VGG19 - OpenCLplaidml: No - Inference - VGG16 - OpenCLcl-mem: Readclpeak: Global Memory Bandwidthshoc: OpenCL - FFT SPshoc: OpenCL - S3Dvkresample: 2x - Singleviennacl: OpenCL BLAS - dGEMM-NTcl-mem: Writeplaidml: No - Inference - ResNet 50 - OpenCLindigobench: OpenCL GPU - Supercarviennacl: OpenCL BLAS - dAXPYbetsy: ETC2 RGB - Highestrealsr-ncnn: 4x - Noviennacl: OpenCL BLAS - dGEMV-Nshoc: OpenCL - Texture Read Bandwidthbetsy: ETC1 - Highestviennacl: OpenCL BLAS - dDOTviennacl: OpenCL BLAS - dCOPYfahbench: viennacl: OpenCL BLAS - dGEMM-TTarrayfire: Conjugate Gradient OpenCLwaifu2x-ncnn: 2x - 3 - Yesviennacl: OpenCL BLAS - sAXPYviennacl: OpenCL BLAS - sDOTviennacl: OpenCL BLAS - sCOPYcl-mem: Copynamd-cuda: ATPase Simulation - 327,506 Atomsviennacl: OpenCL BLAS - dGEMV-Tmixbench: NVIDIA CUDA - Integermixbench: NVIDIA CUDA - Double Precisionmixbench: NVIDIA CUDA - Single PrecisionRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiTITAN RTXRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090299.301767.427122.60143.45545.19183.315308033.93121.5063.852.888311.482611.2430.1561.153.91258.518314.10258.32263.138246.825437.698298.468206.7116219.423023.531083.2824493984571432432864314285718.301012914857147090.647201.013499007.34714.1314480.23242.6832284.511468037676.123493227101.87128.87396.8368.68949.960188.78322.327244333.0378.7523.2993888.48911.6423721110.736.139401368207.99552.0765.322327255280288.20.155023257811.27221.658225.59308.701819.257244.69146.54556.43186.125197853.89122.6564.832.818451.072680.1230.3960.403.91262.868449.25262.90267.898365.015530.458439.958319.6816496.223112.871113.4925195712428572482940517142918.670213146571437155.077362.193555387.40913.2614636.77242.9420454.481473037474.975500673104.17131.82396.8369.26954.251192.70622.297250324.0380.7223.44038811.4933691081.44400367205.30822.0915.305322258270284.60.152473197995.96226.328377.73174.291011.958456.1790.56361.62130.7679310084.9797.7551.853.589791.333017.8625.9051.394.53307.319780.52305.23305.279756.506445.649764.679705.6319106.953269.42595.81278110499142862833395590000021.387915357142868446.248534.724123887.64813.1019471.03258.3637085.441735335164.220580567120.83152.36392.0363.951088.04199.71719.881278314.9418.6824.45138210.1653701144.90397363227.94192.0894.853321258278289.00.143233259219.34256.339668.05169.34975.229258.5886.74354.5129.027809595.1596.8651.523.7110886.653346.8829.0850.154.69341.3610901.17338.65345.8110862.127123.9510871.1810806.2421212.963350.00585.52323123001428573263799864285724.010016967333339079.549468.384528837.74212.1621586.18260.4580525.581866033859.556642760129.95164.37396.8369.301099.39197.20619.053320330.7442.2924.8813886.8739.4933711119.865.017404370244.51843242.0744.648323251278290.20.1380133410234.25288.6610803.63160.42913.2610434.5581.88344.55123.537999815.3795.6850.743.8511917.033667.1725.6844.554.84370.7711920.76371.68378.2511820.677814.8311924.6711765.0923328.093719.30555.03353135967714293544174605714326.4571189361666710392.3310383.045062177.96812.1523388.22270.3021475.772057632254.068712650145.29183.11437.5405.761203.71212.73817.516349347.0472.0925.7044278.8803891168.27441405261.56803501.8974.395330251288301.70.1352235611161.70314.8511622.18152.39899.5913781.3373.30283.9698.6499013246.8372.4238.495.0116487.614821.3819.2433.926.51513.1016498.51513.77522.3116358.3510811.5716457.9116279.2732202.584835.13548.91469178153285714755407435714334.9652246056666713127.4614076.5764481311.03115.2828438.06359.8439047.102773424544.225892410187.25235.75545.6507.081483.30271.24714.781466447.0638.3933.3525187.6294601180.65517477305.31034681.6773.872404319304322.80.1332437514731.97409.0514851.91152.70907.4614490.0073.43281.0897.57100613867.0469.4636.795.2117203.154994.9318.7232.746.92536.0217206.48535.94548.4817030.3911258.4417183.0416941.2833598.085057.65551.44496186879428574865668448571436.7519258308333313853.3914765.3867887511.59216.8328374.42382.4211537.212916823341.913944400197.09248.34567.8530.291578.37288.60013.560493494.7652.4135.3145427.3374621168.30531495307.58604951.6403.855403316305319.30.1330437215399.95425.4515431.56127.15876.6812561.5677.87316.30115.2185211685.3684.5543.853.619039.773490.1623.5653.924.56213.536818.80213.55214.356809.184488.856836.836780.5413718.142770.78962.3120477623285712042400501428615.159111197000006338.905969.652999138.50415.3613379.48293.7797256.061867530067.13442760997.37122.25318.0315.44934.170172.08323.135302.2340.1325.86032710.3622211255.28329315182.00352.4944.850306289275270.00.154713035883.15178.2012815.8385.37552.1116021.6353.66218.4582.07125616587.1961.6731.824.9112854.495040.8717.4239.246.04305.199684.39304.87303.929700.506403.149720.679646.8519484.463432.76596.57291110687428572923415844285721.613015899000008378.398477.2342798911.45120.9919243.59384.1780147.612663923750.635587145133.75168.99393.1388.871153.04211.71717.673293383.8456.4034.0063998.3172242120.83398371236.50582.0834.180359328289293.20.129283338458.84255.2218247.6176.05466.4220072.7246.99189.9172.98141718288.2554.1328.955.8115292.855790.3415.7333.506.98362.4311498.10361.24362.3011545.087629.6711572.6011492.6323200.573795.35507.79338131451714293374028150000025.5521188668571410264.4710070.3250318812.89223.4522837.37413.1817898.432925322746.633687591154.05194.52393.2389.261135.09218.30617.549339380.4495.6837.3663977.8152212131.07398370267.06982.0874.035358326281294.90.1261133310033.78302.0221395.3773.41459.7821524.4944.18184.7667.18148619828.7152.6027.806.0315511.755805.7517.0934.177.41367.8811707.87366.91368.7111695.967730.7511727.9211649.9023509.084710.76501.65345133751285713434147745714326.2725191694285710957.1010205.7651271313.73323.7723025.87454.4966168.673173920140.564720064160.06202.76539.1528.731536.78286.10713.557347516.1513.3738.9475285.7367.0442072.974.249521478267.28191.6683.730435331337.60.1248510120.80305.0421337.8360.87407.3329577.1336.09146.2952.801785237311.6741.4121.948.0022809.117417.8911.2822.799.99536.4517183.85534.50545.9617087.3311290.9817097.7817024.2734278.046305.82442.84499192576285714985896241428637.5181273858571415474.8914570.6871997517.54128.2031733.30568.25721210.793747716432.370991970223.94281.36672.8662.391918.51339.02311.268501646.2730.1646.4966294.3876.0362412210.173.317601555321.33055011.5553.366480367356354.70.1209937713956.21421.8829218.8552.64358.7833723.3431.68124.4345.882049278213.5135.4618.859.2825895.838125.719.7119.9211.54611.0319639.66604.88623.3319501.4712829.1619465.2719408.5639163.488100.68386.72559216602428575576584648571441.8308307015714317317.3116203.0180760020.03733.2535590.07666.47857511.963940914329.2861097120248.34311.73804.8794.552289.54419.3289.561561733.2789.3551.4417054.0825.6352302167.413.084644592335.09815601.5133.283489363358355.00.1206636715545.63469.1632644.6851.28345.1935210.0329.92120.3844.122099285714.0634.4018.369.7227527.948463.769.4819.1111.99651.8920790.49651.06652.3420714.1613650.6620789.4820597.7041672.108327.84373.69600228282500005986997870000044.1102324251666718051.5916881.7684788620.76534.2737668.36679.58418112.354034813928.1801153000257.64323.31825.7813.572346.13429.5999.302602745.5816.8553.0117233.9395.5172219.903.012657607344.98751.4753.235498365363.00.1198816502.90494.2434169.74OpenBenchmarking.org

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: Classroom - Compute: CUDARTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX70140210280350SE +/- 0.03, N = 3SE +/- 0.11, N = 3SE +/- 0.41, N = 3SE +/- 0.20, N = 3SE +/- 0.08, N = 3SE +/- 0.09, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.07, N = 3SE +/- 0.07, N = 3SE +/- 0.10, N = 3SE +/- 0.07, N = 3SE +/- 0.18, N = 3299.30308.70174.29169.34160.42152.39127.1585.3776.0573.4160.8752.6451.28152.70

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Barbershop - Compute: NVIDIA OptiXRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX400800120016002000SE +/- 0.20, N = 3SE +/- 0.26, N = 3SE +/- 1.83, N = 3SE +/- 3.68, N = 3SE +/- 1.18, N = 3SE +/- 0.69, N = 3SE +/- 0.51, N = 3SE +/- 0.26, N = 3SE +/- 0.33, N = 3SE +/- 0.30, N = 3SE +/- 0.86, N = 3SE +/- 0.34, N = 3SE +/- 0.77, N = 3SE +/- 3.15, N = 31767.421819.251011.95975.22913.26899.59876.68552.11466.42459.78407.33358.78345.19907.46

clpeak

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

OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Single-Precision FloatRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX8K16K24K32K40KSE +/- 77.28, N = 15SE +/- 99.77, N = 15SE +/- 77.61, N = 15SE +/- 89.35, N = 15SE +/- 83.80, N = 15SE +/- 222.21, N = 15SE +/- 16.58, N = 14SE +/- 7.12, N = 15SE +/- 12.54, N = 14SE +/- 2.30, N = 14SE +/- 69.64, N = 13SE +/- 21.75, N = 15SE +/- 28.06, N = 14SE +/- 183.70, N = 157122.607244.698456.179258.5810434.5513781.3312561.5616021.6320072.7221524.4929577.1333723.3435210.0314490.001. (CXX) g++ options: -O3 -rdynamic -lOpenCL

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: Classroom - Compute: NVIDIA OptiXRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX306090120150SE +/- 0.27, N = 3SE +/- 0.63, N = 3SE +/- 0.02, N = 3SE +/- 0.12, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.22, N = 3SE +/- 0.23, N = 3SE +/- 0.06, N = 3SE +/- 0.31, N = 3SE +/- 0.16, N = 3143.45146.5490.5686.7481.8873.3077.8753.6646.9944.1836.0931.6829.9273.43

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Pabellon Barcelona - Compute: CUDARTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX120240360480600SE +/- 0.11, N = 3SE +/- 0.25, N = 3SE +/- 0.34, N = 3SE +/- 0.07, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.09, N = 3SE +/- 0.13, N = 3SE +/- 0.03, N = 3SE +/- 0.09, N = 3545.19556.43361.62354.50344.55283.96316.30218.45189.91184.76146.29124.43120.38281.08

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Pabellon Barcelona - Compute: NVIDIA OptiXRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX4080120160200SE +/- 0.57, N = 3SE +/- 0.54, N = 3SE +/- 0.02, N = 3SE +/- 0.14, N = 3SE +/- 0.03, N = 3SE +/- 0.06, N = 3SE +/- 0.09, N = 3SE +/- 0.09, N = 3SE +/- 0.07, N = 3SE +/- 0.08, N = 3SE +/- 0.14, N = 3SE +/- 0.03, N = 3SE +/- 0.08, N = 3SE +/- 0.07, N = 3183.31186.12130.76129.02123.5398.64115.2182.0772.9867.1852.8045.8844.1297.57

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.orgvpaths, More Is BetterChaos Group V-RAY 5Mode: NVIDIA CUDA GPURTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX5001000150020002500SE +/- 0.33, N = 3SE +/- 1.15, N = 3SE +/- 0.58, N = 3SE +/- 0.67, N = 3SE +/- 0.58, N = 3SE +/- 0.88, N = 3SE +/- 2.03, N = 3SE +/- 1.33, N = 3SE +/- 0.88, N = 35305197937807999908521256141714861785204920991006

OpenBenchmarking.orgvrays, More Is BetterChaos Group V-RAY 5Mode: NVIDIA RTX GPURTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX6001200180024003000SE +/- 0.88, N = 3SE +/- 0.67, N = 3SE +/- 0.67, N = 3SE +/- 0.67, N = 3SE +/- 1.33, N = 3SE +/- 1.76, N = 3SE +/- 0.67, N = 3SE +/- 0.67, N = 3SE +/- 3.00, N = 3SE +/- 3.48, N = 3SE +/- 9.02, N = 3SE +/- 1.76, N = 38037851008959981132411681658182819822373278228571386

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: DLSC - Acceleration: GPURTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX48121620SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 33.933.894.975.155.376.835.367.198.258.7111.6713.5114.067.04MIN: 3.81 / MAX: 4.12MIN: 3.78 / MAX: 4.04MIN: 4.86 / MAX: 5.09MIN: 5.03 / MAX: 5.33MIN: 5.07 / MAX: 5.51MIN: 6.56 / MAX: 6.93MIN: 5.23 / MAX: 5.58MIN: 6.89 / MAX: 7.44MIN: 8.13 / MAX: 8.41MIN: 8.19 / MAX: 8.87MIN: 11.22 / MAX: 11.86MIN: 12.39 / MAX: 13.74MIN: 13.57 / MAX: 14.34MIN: 6.59 / MAX: 7.3

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: Fishy Cat - Compute: CUDARTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX306090120150SE +/- 0.18, N = 3SE +/- 0.07, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3121.50122.6597.7596.8695.6872.4284.5561.6754.1352.6041.4135.4634.4069.46

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: BMW27 - Compute: CUDARTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX1428425670SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 363.8564.8351.8551.5250.7438.4943.8531.8228.9527.8021.9418.8518.3636.79

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX3691215SE +/- 0.03, N = 4SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.04, N = 6SE +/- 0.07, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 3SE +/- 0.02, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 3SE +/- 0.05, N = 32.882.813.583.713.855.013.614.915.816.038.009.289.725.21MIN: 0.92 / MAX: 3.41MIN: 0.88 / MAX: 3.32MIN: 1.17 / MAX: 4.31MIN: 1.18 / MAX: 4.47MIN: 1.19 / MAX: 4.64MIN: 1.68 / MAX: 6.06MIN: 1.15 / MAX: 4.3MIN: 1.66 / MAX: 5.96MIN: 1.82 / MAX: 7.01MIN: 2.03 / MAX: 7.29MIN: 2.38 / MAX: 9.7MIN: 2.69 / MAX: 11.32MIN: 3.01 / MAX: 11.82MIN: 1.69 / MAX: 6.28

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-vec4RTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX6K12K18K24K30KSE +/- 42.90, N = 3SE +/- 46.44, N = 3SE +/- 43.80, N = 3SE +/- 59.30, N = 3SE +/- 60.73, N = 3SE +/- 79.54, N = 3SE +/- 12.40, N = 3SE +/- 31.64, N = 3SE +/- 58.84, N = 3SE +/- 78.89, N = 3SE +/- 75.50, N = 3SE +/- 108.44, N = 3SE +/- 137.76, N = 3SE +/- 134.86, N = 38311.488451.079791.3310886.6511917.0316487.619039.7712854.4915292.8515511.7522809.1125895.8327527.9417203.15

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.orgGFLOPS, More Is BetterArrayFire 3.7Test: BLAS OpenCLRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX2K4K6K8K10KSE +/- 0.91, N = 4SE +/- 6.61, N = 4SE +/- 0.43, N = 4SE +/- 7.17, N = 4SE +/- 0.34, N = 4SE +/- 8.53, N = 4SE +/- 6.19, N = 4SE +/- 0.26, N = 4SE +/- 0.72, N = 4SE +/- 17.44, N = 4SE +/- 5.63, N = 4SE +/- 1.74, N = 4SE +/- 3.92, N = 4SE +/- 12.10, N = 42611.242680.123017.863346.883667.174821.383490.165040.875790.345805.757417.898125.718463.764994.931. (CXX) g++ options: -rdynamic

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: BMW27 - Compute: NVIDIA OptiXRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX714212835SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 2.65, N = 15SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 4SE +/- 2.65, N = 15SE +/- 0.01, N = 4SE +/- 0.01, N = 5SE +/- 0.01, N = 5SE +/- 0.01, N = 330.1530.3925.9029.0825.6819.2423.5617.4215.7317.0911.289.719.4818.72

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Fishy Cat - Compute: NVIDIA OptiXRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX1428425670SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 2.59, N = 15SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.07, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 2.61, N = 15SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 361.1560.4051.3950.1544.5533.9253.9239.2433.5034.1722.7919.9219.1132.74

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: LuxCore Benchmark - Acceleration: GPURTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX3691215SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 33.913.914.534.694.846.514.566.046.987.419.9911.5411.996.92MIN: 1.4 / MAX: 4.62MIN: 1.41 / MAX: 4.63MIN: 1.3 / MAX: 5.39MIN: 1.66 / MAX: 5.58MIN: 1.66 / MAX: 5.75MIN: 1.93 / MAX: 7.74MIN: 1.64 / MAX: 5.37MIN: 1.87 / MAX: 7.15MIN: 2.27 / MAX: 8.27MIN: 2.28 / MAX: 8.81MIN: 3.32 / MAX: 11.84MIN: 3.63 / MAX: 13.73MIN: 3.93 / MAX: 14.25MIN: 2.23 / MAX: 8.19

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 20210424fp64-vec4RTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX140280420560700SE +/- 0.65, N = 3SE +/- 0.06, N = 2SE +/- 0.50, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 2SE +/- 1.31, N = 3SE +/- 0.03, N = 3SE +/- 0.31, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.35, N = 3SE +/- 0.61, N = 3SE +/- 0.79, N = 3SE +/- 1.43, N = 3258.51262.86307.31341.36370.77513.10213.53305.19362.43367.88536.45611.03651.89536.02

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp32-scalarRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX4K8K12K16K20KSE +/- 61.71, N = 3SE +/- 56.83, N = 3SE +/- 43.99, N = 3SE +/- 88.42, N = 3SE +/- 53.05, N = 3SE +/- 114.67, N = 3SE +/- 0.24, N = 3SE +/- 13.93, N = 3SE +/- 7.52, N = 3SE +/- 49.99, N = 3SE +/- 74.66, N = 3SE +/- 104.27, N = 3SE +/- 65.40, N = 3SE +/- 155.55, N = 38314.108449.259780.5210901.1711920.7616498.516818.809684.3911498.1011707.8717183.8519639.6620790.4917206.48

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp64-scalarRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX140280420560700SE +/- 0.15, N = 3SE +/- 0.03, N = 3SE +/- 0.79, N = 3SE +/- 0.02, N = 3SE +/- 0.95, N = 3SE +/- 1.54, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.34, N = 3SE +/- 0.95, N = 3SE +/- 1.32, N = 3SE +/- 0.01, N = 3SE +/- 1.37, N = 3SE +/- 1.43, N = 3258.32262.90305.23338.65371.68513.77213.55304.87361.24366.91534.50604.88651.06535.94

clpeak

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

OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Double-Precision DoubleRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX140280420560700SE +/- 0.74, N = 3SE +/- 0.74, N = 3SE +/- 0.80, N = 3SE +/- 0.84, N = 3SE +/- 0.96, N = 3SE +/- 1.40, N = 3SE +/- 0.01, N = 3SE +/- 0.38, N = 3SE +/- 0.02, N = 3SE +/- 0.84, N = 3SE +/- 1.37, N = 3SE +/- 1.38, N = 3SE +/- 0.05, N = 3SE +/- 1.27, N = 3263.13267.89305.27345.81378.25522.31214.35303.92362.30368.71545.96623.33652.34548.481. (CXX) g++ options: -O3 -rdynamic -lOpenCL

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 20210424int32-scalarRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX4K8K12K16K20KSE +/- 21.65, N = 3SE +/- 0.21, N = 3SE +/- 8.24, N = 3SE +/- 1.08, N = 3SE +/- 30.36, N = 3SE +/- 43.31, N = 3SE +/- 0.77, N = 3SE +/- 8.15, N = 3SE +/- 0.75, N = 3SE +/- 0.71, N = 3SE +/- 1.14, N = 3SE +/- 0.24, N = 3SE +/- 26.66, N = 3SE +/- 44.08, N = 38246.828365.019756.5010862.1211820.6716358.356809.189700.5011545.0811695.9617087.3319501.4720714.1617030.39

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int16-scalarRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX3K6K9K12K15KSE +/- 0.09, N = 3SE +/- 0.15, N = 3SE +/- 0.30, N = 3SE +/- 0.36, N = 3SE +/- 20.31, N = 3SE +/- 27.43, N = 3SE +/- 5.77, N = 3SE +/- 3.46, N = 3SE +/- 0.14, N = 3SE +/- 0.05, N = 3SE +/- 1.14, N = 3SE +/- 20.56, N = 3SE +/- 1.11, N = 3SE +/- 27.84, N = 35437.695530.456445.647123.957814.8310811.574488.856403.147629.677730.7511290.9812829.1613650.6611258.44

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp16-scalarRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX4K8K12K16K20KSE +/- 21.01, N = 3SE +/- 22.92, N = 3SE +/- 32.23, N = 3SE +/- 29.14, N = 3SE +/- 60.69, N = 3SE +/- 43.12, N = 3SE +/- 9.15, N = 3SE +/- 25.27, N = 3SE +/- 43.84, N = 3SE +/- 57.85, N = 3SE +/- 43.04, N = 3SE +/- 30.73, N = 3SE +/- 80.41, N = 3SE +/- 92.50, N = 38298.468439.959764.6710871.1811924.6716457.916836.839720.6711572.6011727.9217097.7819465.2720789.4817183.04

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int32-vec4RTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX4K8K12K16K20KSE +/- 20.82, N = 3SE +/- 0.63, N = 3SE +/- 8.84, N = 3SE +/- 0.27, N = 3SE +/- 30.69, N = 3SE +/- 45.18, N = 3SE +/- 0.54, N = 3SE +/- 4.72, N = 3SE +/- 1.00, N = 3SE +/- 1.12, N = 3SE +/- 0.02, N = 3SE +/- 0.25, N = 3SE +/- 2.62, N = 3SE +/- 46.17, N = 38206.718319.689705.6310806.2411765.0916279.276780.549646.8511492.6311649.9017024.2719408.5620597.7016941.28

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp16-vec4RTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX9K18K27K36K45KSE +/- 40.81, N = 3SE +/- 45.05, N = 3SE +/- 64.46, N = 3SE +/- 0.93, N = 3SE +/- 118.62, N = 3SE +/- 86.26, N = 3SE +/- 0.44, N = 3SE +/- 49.62, N = 3SE +/- 88.97, N = 3SE +/- 114.74, N = 3SE +/- 88.40, N = 3SE +/- 51.02, N = 3SE +/- 91.14, N = 3SE +/- 148.11, N = 316219.4216496.2219106.9521212.9623328.0932202.5813718.1419484.4623200.5723509.0834278.0439163.4841672.1033598.08

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: GEMM SGEMM_NRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX2K4K6K8K10KSE +/- 7.64, N = 9SE +/- 9.39, N = 9SE +/- 0.86, N = 9SE +/- 1.20, N = 9SE +/- 1.82, N = 10SE +/- 27.28, N = 10SE +/- 5.34, N = 9SE +/- 2.76, N = 10SE +/- 20.27, N = 10SE +/- 5.86, N = 10SE +/- 18.49, N = 11SE +/- 30.90, N = 12SE +/- 31.35, N = 12SE +/- 26.53, N = 103023.533112.873269.423350.003719.304835.132770.783432.763795.354710.766305.828100.688327.845057.651. (CXX) g++ options: -O2 -lSHOCCommonMPI -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt -pthread -lmpi_cxx -lmpi

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: CUDARTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX2004006008001000SE +/- 0.20, N = 3SE +/- 0.11, N = 3SE +/- 0.01, N = 3SE +/- 0.19, N = 3SE +/- 0.07, N = 3SE +/- 0.25, N = 3SE +/- 0.24, N = 3SE +/- 0.31, N = 3SE +/- 0.24, N = 3SE +/- 0.27, N = 3SE +/- 0.45, N = 3SE +/- 0.21, N = 3SE +/- 0.32, N = 3SE +/- 0.26, N = 31083.281113.49595.81585.52555.03548.91962.31596.57507.79501.65442.84386.72373.69551.44

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-NNRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX130260390520650SE +/- 1.20, N = 3SE +/- 1.00, N = 3SE +/- 0.67, N = 3SE +/- 0.67, N = 3SE +/- 0.67, N = 3SE +/- 1.20, N = 3SE +/- 1.33, N = 3SE +/- 1.20, N = 32442512783233534692042913383454995596004961. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

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: SHA1RTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX5000M10000M15000M20000M25000MSE +/- 9752578.59, N = 7SE +/- 13731990.06, N = 7SE +/- 5146308.37, N = 7SE +/- 25963370.33, N = 7SE +/- 15457664.37, N = 7SE +/- 22447650.36, N = 7SE +/- 7946598.98, N = 7SE +/- 5748658.52, N = 7SE +/- 23586201.72, N = 7SE +/- 6164287.09, N = 7SE +/- 23290062.40, N = 7SE +/- 27258234.73, N = 7SE +/- 26924236.29, N = 6SE +/- 22120437.19, N = 79398457143957124285711049914286123001428571359677142917815328571776232857111068742857131451714291337512857119257628571216602428572282825000018687942857

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX130260390520650SE +/- 3.00, N = 2SE +/- 0.67, N = 3SE +/- 0.67, N = 3SE +/- 1.33, N = 3SE +/- 1.33, N = 3SE +/- 1.20, N = 32432482833263544752042903373434985575984861. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

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: MD5RTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX15000M30000M45000M60000M75000MSE +/- 49115005.26, N = 7SE +/- 62259505.74, N = 7SE +/- 40022380.64, N = 7SE +/- 104693541.11, N = 7SE +/- 48387138.48, N = 7SE +/- 73945478.17, N = 7SE +/- 33928239.60, N = 7SE +/- 50524660.91, N = 7SE +/- 64426488.19, N = 7SE +/- 82415641.88, N = 7SE +/- 119086542.02, N = 7SE +/- 114085221.76, N = 7SE +/- 104099604.32, N = 7SE +/- 90062008.87, N = 72864314285729405171429339559000003799864285741746057143540743571432400501428634158442857402815000004147745714358962414286658464857146997870000056684485714

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.orgGHash/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2020-04-17Target: OpenCL - Benchmark: MD5 HashRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX1020304050SE +/- 0.01, N = 15SE +/- 0.01, N = 13SE +/- 0.01, N = 13SE +/- 0.01, N = 15SE +/- 0.00, N = 13SE +/- 0.04, N = 14SE +/- 0.01, N = 12SE +/- 0.03, N = 13SE +/- 0.05, N = 13SE +/- 0.01, N = 13SE +/- 0.00, N = 14SE +/- 0.06, N = 14SE +/- 0.04, N = 14SE +/- 0.04, N = 1418.3018.6721.3924.0126.4634.9715.1621.6125.5526.2737.5241.8344.1136.751. (CXX) g++ options: -O2 -lSHOCCommonMPI -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt -pthread -lmpi_cxx -lmpi

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: SHA-512RTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX700M1400M2100M2800M3500MSE +/- 880321.52, N = 7SE +/- 1259413.53, N = 7SE +/- 430076.72, N = 7SE +/- 2564067.95, N = 6SE +/- 1633078.21, N = 6SE +/- 2857349.20, N = 6SE +/- 1910248.05, N = 7SE +/- 2343684.52, N = 7SE +/- 2159742.99, N = 7SE +/- 2485372.17, N = 7SE +/- 3881545.38, N = 7SE +/- 2909578.36, N = 7SE +/- 2166012.72, N = 6SE +/- 3167166.63, N = 612914857141314657143153571428616967333331893616667246056666711197000001589900000188668571419169428572738585714307015714332425166672583083333

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX4K8K12K16K20KSE +/- 85.91, N = 15SE +/- 84.44, N = 15SE +/- 83.15, N = 15SE +/- 73.78, N = 15SE +/- 83.89, N = 15SE +/- 52.71, N = 12SE +/- 3.88, N = 15SE +/- 76.78, N = 15SE +/- 62.48, N = 13SE +/- 52.74, N = 13SE +/- 110.02, N = 15SE +/- 114.00, N = 12SE +/- 105.32, N = 12SE +/- 163.86, N = 157090.647155.078446.249079.5410392.3313127.466338.908378.3910264.4710957.1015474.8917317.3118051.5913853.391. (CXX) g++ options: -O3 -rdynamic -lOpenCL

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX4K8K12K16K20KSE +/- 14.82, N = 3SE +/- 0.29, N = 3SE +/- 21.85, N = 3SE +/- 1.82, N = 3SE +/- 0.06, N = 3SE +/- 14.70, N = 3SE +/- 7.82, N = 3SE +/- 11.39, N = 3SE +/- 0.82, N = 3SE +/- 0.81, N = 3SE +/- 0.42, N = 3SE +/- 4.40, N = 3SE +/- 4.03, N = 3SE +/- 5.90, N = 37201.017362.198534.729468.3810383.0414076.575969.658477.2310070.3210205.7614570.6816203.0116881.7614765.38

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX200K400K600K800K1000KSE +/- 208.74, N = 8SE +/- 350.48, N = 8SE +/- 432.37, N = 8SE +/- 841.20, N = 6SE +/- 4311.72, N = 6SE +/- 2086.05, N = 8SE +/- 121.65, N = 8SE +/- 78.96, N = 9SE +/- 863.64, N = 8SE +/- 463.85, N = 8SE +/- 816.73, N = 8SE +/- 528.10, N = 9SE +/- 405.57, N = 7SE +/- 340.04, N = 8349900355538412388452883506217644813299913427989503188512713719975807600847886678875

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX510152025SE +/- 0.016, N = 3SE +/- 0.001, N = 3SE +/- 0.002, N = 3SE +/- 0.005, N = 3SE +/- 0.005, N = 3SE +/- 0.004, N = 3SE +/- 0.006, N = 3SE +/- 0.004, N = 3SE +/- 0.001, N = 3SE +/- 0.001, N = 3SE +/- 0.004, N = 3SE +/- 0.005, N = 3SE +/- 0.003, N = 3SE +/- 0.003, N = 37.3477.4097.6487.7427.96811.0318.50411.45112.89213.73317.54120.03720.76511.592

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: Rainbow Colors and Prism - Acceleration: GPURTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX816243240SE +/- 0.01, N = 4SE +/- 0.01, N = 4SE +/- 0.03, N = 4SE +/- 0.00, N = 4SE +/- 0.01, N = 4SE +/- 0.02, N = 4SE +/- 0.00, N = 4SE +/- 0.05, N = 5SE +/- 0.05, N = 6SE +/- 0.06, N = 6SE +/- 0.08, N = 6SE +/- 0.24, N = 6SE +/- 0.00, N = 6SE +/- 0.01, N = 514.1313.2613.1012.1612.1515.2815.3620.9923.4523.7728.2033.2534.2716.83MIN: 13.33 / MAX: 14.65MIN: 12.41 / MAX: 14.02MIN: 12.42 / MAX: 13.77MIN: 11.24 / MAX: 12.82MIN: 11.25 / MAX: 12.81MIN: 13.65 / MAX: 16.66MIN: 13.94 / MAX: 16.32MIN: 19.21 / MAX: 22.15MIN: 20.97 / MAX: 24.32MIN: 21.81 / MAX: 25.33MIN: 26.09 / MAX: 31.07MIN: 30.01 / MAX: 35.85MIN: 30.52 / MAX: 36.32MIN: 15.76 / MAX: 18.22

Mixbench

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

OpenBenchmarking.orgGFLOPS, More Is BetterMixbench 2020-06-23Backend: NVIDIA CUDA - Benchmark: Half PrecisionRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX8K16K24K32K40KSE +/- 52.79, N = 14SE +/- 42.04, N = 14SE +/- 45.34, N = 14SE +/- 33.25, N = 15SE +/- 70.30, N = 14SE +/- 105.21, N = 14SE +/- 0.78, N = 14SE +/- 1.61, N = 14SE +/- 2.92, N = 14SE +/- 4.24, N = 14SE +/- 632.41, N = 15SE +/- 685.00, N = 15SE +/- 800.73, N = 15SE +/- 394.62, N = 1514480.2314636.7719471.0321586.1823388.2228438.0613379.4819243.5922837.3723025.8731733.3035590.0737668.3628374.421. (CXX) g++ options: -lm -lstdc++ -lOpenCL -lrt -O2

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX150300450600750242.68242.94258.36260.46270.30359.84293.78384.18413.18454.50568.26666.48679.58382.42

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: Orange Juice - Acceleration: GPURTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX3691215SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.07, N = 3SE +/- 0.01, N = 3SE +/- 0.06, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 34.514.485.445.585.777.106.067.618.438.6710.7911.9612.357.21MIN: 3.75 / MAX: 4.99MIN: 3.75 / MAX: 4.81MIN: 4.44 / MAX: 6.12MIN: 4.54 / MAX: 6.31MIN: 4.67 / MAX: 6.58MIN: 5.6 / MAX: 8.83MIN: 4.73 / MAX: 7.39MIN: 6.03 / MAX: 9.24MIN: 6.69 / MAX: 10.55MIN: 6.96 / MAX: 10.98MIN: 8.72 / MAX: 14.46MIN: 9.91 / MAX: 16.43MIN: 10.23 / MAX: 17.13MIN: 5.68 / MAX: 9.07

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX9K18K27K36K45KSE +/- 72.46, N = 3SE +/- 32.62, N = 3SE +/- 217.96, N = 3SE +/- 29.46, N = 3SE +/- 110.32, N = 3SE +/- 103.74, N = 3SE +/- 110.08, N = 3SE +/- 77.49, N = 3SE +/- 168.57, N = 3SE +/- 101.51, N = 3SE +/- 186.45, N = 3SE +/- 106.84, N = 3SE +/- 52.75, N = 3SE +/- 46.38, N = 314680147301735318660205762773418675266392925331739374773940940348291681. (CXX) g++ options: -flto -pthread

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX80160240320400SE +/- 0.88, N = 3SE +/- 1.33, N = 3SE +/- 1.67, N = 3SE +/- 0.88, N = 3SE +/- 1.20, N = 3SE +/- 1.20, N = 3SE +/- 1.20, N = 3SE +/- 0.88, N = 3SE +/- 1.00, N = 3SE +/- 0.67, N = 3SE +/- 0.67, N = 3SE +/- 0.58, N = 3SE +/- 0.88, N = 3376374351338322245300237227201164143139233

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: YesRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX20406080100SE +/- 0.31, N = 3SE +/- 0.43, N = 3SE +/- 0.22, N = 3SE +/- 0.30, N = 3SE +/- 0.14, N = 3SE +/- 0.17, N = 3SE +/- 0.01, N = 3SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.08, N = 3SE +/- 0.09, N = 3SE +/- 0.08, N = 3SE +/- 0.06, N = 3SE +/- 0.18, N = 376.1274.9864.2259.5654.0744.2367.1350.6446.6340.5632.3729.2928.1841.91

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: 7-ZipRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX200K400K600K800K1000KSE +/- 469.64, N = 11SE +/- 338.47, N = 11SE +/- 1961.31, N = 15SE +/- 447.26, N = 10SE +/- 498.27, N = 10SE +/- 337.46, N = 10SE +/- 258.83, N = 11SE +/- 305.81, N = 11SE +/- 535.21, N = 11SE +/- 340.13, N = 11SE +/- 415.55, N = 10SE +/- 815.38, N = 10SE +/- 879.27, N = 10SE +/- 555.18, N = 1049322750067358056764276071265089241042760958714568759172006499197010971201153000944400

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX60120180240300SE +/- 0.28, N = 3SE +/- 0.39, N = 3SE +/- 0.17, N = 3SE +/- 0.58, N = 3SE +/- 0.32, N = 3SE +/- 0.55, N = 3SE +/- 0.06, N = 3SE +/- 0.33, N = 3SE +/- 0.31, N = 3SE +/- 0.28, N = 3SE +/- 0.38, N = 3SE +/- 0.52, N = 3SE +/- 0.33, N = 3SE +/- 0.48, N = 3101.87104.17120.83129.95145.29187.2597.37133.75154.05160.06223.94248.34257.64197.09

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: VGG16 - Device: OpenCLRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX70140210280350SE +/- 0.41, N = 3SE +/- 0.54, N = 3SE +/- 0.34, N = 3SE +/- 0.75, N = 3SE +/- 0.45, N = 3SE +/- 0.63, N = 3SE +/- 0.09, N = 3SE +/- 0.35, N = 3SE +/- 0.45, N = 3SE +/- 0.72, N = 3SE +/- 0.47, N = 4SE +/- 0.58, N = 4SE +/- 0.49, N = 4SE +/- 0.54, N = 3128.87131.82152.36164.37183.11235.75122.25168.99194.52202.76281.36311.73323.31248.34

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: ReadRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX2004006008001000SE +/- 0.10, N = 8SE +/- 0.10, N = 8SE +/- 0.06, N = 8SE +/- 0.08, N = 8SE +/- 0.11, N = 8SE +/- 0.15, N = 8SE +/- 0.11, N = 7SE +/- 0.18, N = 8SE +/- 0.19, N = 8SE +/- 0.11, N = 9SE +/- 0.18, N = 9SE +/- 0.37, N = 9SE +/- 0.18, N = 10SE +/- 0.09, N = 9396.8396.8392.0396.8437.5545.6318.0393.1393.2539.1672.8804.8825.7567.81. (CC) gcc options: -O2 -flto -lOpenCL

clpeak

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

OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Global Memory BandwidthRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX2004006008001000SE +/- 0.08, N = 13SE +/- 0.05, N = 15SE +/- 0.07, N = 13SE +/- 0.14, N = 15SE +/- 0.21, N = 13SE +/- 0.45, N = 12SE +/- 0.01, N = 11SE +/- 0.22, N = 13SE +/- 0.25, N = 13SE +/- 0.40, N = 13SE +/- 0.07, N = 12SE +/- 0.03, N = 13SE +/- 0.01, N = 14SE +/- 0.03, N = 13368.68369.26363.95369.30405.76507.08315.44388.87389.26528.73662.39794.55813.57530.291. (CXX) g++ options: -O3 -rdynamic -lOpenCL

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: FFT SPRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX5001000150020002500SE +/- 0.12, N = 12SE +/- 0.42, N = 12SE +/- 0.30, N = 12SE +/- 0.39, N = 12SE +/- 0.30, N = 12SE +/- 1.41, N = 12SE +/- 0.06, N = 12SE +/- 0.15, N = 12SE +/- 0.32, N = 12SE +/- 0.35, N = 12SE +/- 0.32, N = 13SE +/- 1.82, N = 13SE +/- 1.60, N = 13SE +/- 0.84, N = 12949.96954.251088.041099.391203.711483.30934.171153.041135.091536.781918.512289.542346.131578.371. (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: S3DRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX90180270360450SE +/- 0.04, N = 13SE +/- 0.07, N = 13SE +/- 0.09, N = 13SE +/- 0.09, N = 12SE +/- 0.07, N = 13SE +/- 0.13, N = 13SE +/- 0.07, N = 13SE +/- 0.10, N = 13SE +/- 0.16, N = 13SE +/- 0.05, N = 12SE +/- 0.19, N = 13SE +/- 0.22, N = 13SE +/- 0.12, N = 13SE +/- 0.08, N = 13188.78192.71199.72197.21212.74271.25172.08211.72218.31286.11339.02419.33429.60288.601. (CXX) g++ options: -O2 -lSHOCCommonMPI -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt -pthread -lmpi_cxx -lmpi

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX612182430SE +/- 0.032, N = 5SE +/- 0.033, N = 5SE +/- 0.018, N = 5SE +/- 0.023, N = 5SE +/- 0.012, N = 5SE +/- 0.020, N = 5SE +/- 0.001, N = 5SE +/- 0.006, N = 5SE +/- 0.004, N = 5SE +/- 0.003, N = 5SE +/- 0.006, N = 5SE +/- 0.007, N = 5SE +/- 0.010, N = 5SE +/- 0.032, N = 522.32722.29719.88119.05317.51614.78123.13517.67317.54913.55711.2689.5619.30213.5601. (CXX) g++ options: -O3 -pthread

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-NTRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX130260390520650SE +/- 2.00, N = 2SE +/- 1.67, N = 3SE +/- 1.33, N = 3SE +/- 1.67, N = 3SE +/- 1.33, N = 32432502783203494662933393475015606024931. (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: WriteRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX160320480640800SE +/- 0.37, N = 8SE +/- 0.35, N = 8SE +/- 0.15, N = 8SE +/- 0.55, N = 8SE +/- 0.31, N = 8SE +/- 0.38, N = 8SE +/- 0.12, N = 7SE +/- 0.07, N = 8SE +/- 0.10, N = 8SE +/- 0.26, N = 9SE +/- 0.35, N = 9SE +/- 0.48, N = 9SE +/- 0.18, N = 10SE +/- 0.75, N = 9333.0324.0314.9330.7347.0447.0302.2383.8380.4516.1646.2733.2745.5494.71. (CC) gcc options: -O2 -flto -lOpenCL

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: ResNet 50 - Device: OpenCLRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX2004006008001000SE +/- 0.43, N = 5SE +/- 0.55, N = 5SE +/- 0.41, N = 5SE +/- 0.92, N = 5SE +/- 0.45, N = 5SE +/- 0.24, N = 6SE +/- 0.20, N = 5SE +/- 0.18, N = 5SE +/- 0.46, N = 6SE +/- 0.66, N = 5SE +/- 0.79, N = 6SE +/- 0.74, N = 6SE +/- 1.07, N = 6SE +/- 0.45, N = 6378.75380.72418.68442.29472.09638.39340.13456.40495.68513.37730.16789.35816.85652.41

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: SupercarRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX1224364860SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.07, N = 3SE +/- 0.08, N = 3SE +/- 0.02, N = 3SE +/- 0.12, N = 3SE +/- 0.03, N = 323.3023.4424.4524.8825.7033.3525.8634.0137.3738.9546.5051.4453.0135.31

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: OpenCL BLAS - dAXPYRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX1603204806408003883883823884275183273993975286297057235421. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

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 2060 SUPERRTX 2080RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090246810SE +/- 0.018, N = 5SE +/- 0.012, N = 6SE +/- 0.011, N = 7SE +/- 0.008, N = 8SE +/- 0.009, N = 8SE +/- 0.012, N = 88.4896.8735.7364.3874.0823.9391. (CXX) g++ options: -O3 -O2 -lpthread -ldl

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX3691215SE +/- 0.012, N = 4SE +/- 0.017, N = 5SE +/- 0.004, N = 5SE +/- 0.020, N = 5SE +/- 0.014, N = 5SE +/- 0.008, N = 6SE +/- 0.009, N = 5SE +/- 0.004, N = 6SE +/- 0.009, N = 6SE +/- 0.008, N = 6SE +/- 0.012, N = 7SE +/- 0.004, N = 7SE +/- 0.005, N = 7SE +/- 0.014, N = 611.64211.49310.1659.4938.8807.62910.3628.3177.8157.0446.0365.6355.5177.337

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: OpenCL BLAS - dGEMV-NRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3080RTX 3080 TiTITAN RTX100200300400500SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 33723693703713894602212242212412304621. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

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.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2020-04-17Target: OpenCL - Benchmark: Texture Read BandwidthRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX5001000150020002500SE +/- 2.79, N = 4SE +/- 3.17, N = 3SE +/- 0.65, N = 3SE +/- 1.56, N = 3SE +/- 2.81, N = 3SE +/- 1.11, N = 3SE +/- 3.60, N = 4SE +/- 2.95, N = 4SE +/- 1.91, N = 4SE +/- 1.31, N = 4SE +/- 0.73, N = 3SE +/- 6.37, N = 3SE +/- 4.61, N = 3SE +/- 3.30, N = 31110.731081.441144.901119.861168.271180.651255.282120.832131.072072.972210.172167.412219.901168.301. (CXX) g++ options: -O2 -lSHOCCommonMPI -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt -pthread -lmpi_cxx -lmpi

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: ETC1 - Quality: HighestRTX 2060 SUPERRTX 2080RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090246810SE +/- 0.009, N = 7SE +/- 0.011, N = 7SE +/- 0.014, N = 8SE +/- 0.006, N = 9SE +/- 0.012, N = 9SE +/- 0.006, N = 96.1395.0174.2493.3173.0843.0121. (CXX) g++ options: -O3 -O2 -lpthread -ldl

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: OpenCL BLAS - dDOTRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX140280420560700SE +/- 0.33, N = 3SE +/- 0.33, N = 34014003974044415173293983985216016446575311. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - dCOPYRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX130260390520650SE +/- 0.33, N = 3SE +/- 0.58, N = 33683673633704054773153713704785555926074951. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX70140210280350SE +/- 0.40, N = 3SE +/- 0.59, N = 3SE +/- 0.18, N = 3SE +/- 0.35, N = 3SE +/- 0.45, N = 3SE +/- 0.75, N = 3SE +/- 0.16, N = 3SE +/- 0.27, N = 3SE +/- 0.47, N = 3SE +/- 0.12, N = 3SE +/- 0.26, N = 3SE +/- 0.79, N = 3SE +/- 0.19, N = 3SE +/- 0.59, N = 3208.00205.31227.94244.52261.57305.31182.00236.51267.07267.28321.33335.10344.99307.59

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 2080RTX 2080 SUPERRTX 2080 TiRTX 3080RTX 3080 TiTITAN RTX120240360480600SE +/- 2.00, N = 23243504685015604951. (CXX) g++ options: -fopenmp -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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX0.56121.12241.68362.24482.806SE +/- 0.002, N = 10SE +/- 0.002, N = 10SE +/- 0.001, N = 10SE +/- 0.001, N = 9SE +/- 0.001, N = 10SE +/- 0.004, N = 11SE +/- 0.001, N = 10SE +/- 0.001, N = 10SE +/- 0.001, N = 10SE +/- 0.001, N = 10SE +/- 0.002, N = 10SE +/- 0.001, N = 10SE +/- 0.000, N = 10SE +/- 0.005, N = 112.0762.0912.0892.0741.8971.6772.4942.0832.0871.6681.5551.5131.4751.6401. (CXX) g++ options: -rdynamic

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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX1.19752.3953.59254.795.9875SE +/- 0.007, N = 7SE +/- 0.008, N = 7SE +/- 0.004, N = 7SE +/- 0.005, N = 8SE +/- 0.006, N = 8SE +/- 0.006, N = 8SE +/- 0.003, N = 7SE +/- 0.003, N = 8SE +/- 0.003, N = 8SE +/- 0.006, N = 9SE +/- 0.004, N = 9SE +/- 0.001, N = 9SE +/- 0.003, N = 9SE +/- 0.008, N = 85.3225.3054.8534.6484.3953.8724.8504.1804.0353.7303.3663.2833.2353.855

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: OpenCL BLAS - sAXPYRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX110220330440550SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.58, N = 33273223213233304043063593584354804894984031. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - sDOTRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3080RTX 3080 TiTITAN RTX80160240320400SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.67, N = 32552572582512513192893273263673633161. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGB/s, More Is BetterViennaCL 1.7.1Test: OpenCL BLAS - sCOPYRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX80160240320400SE +/- 1.20, N = 3SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 0.88, N = 3SE +/- 1.33, N = 3SE +/- 1.45, N = 32772682782782883042752882803313563563653051. (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 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX80160240320400SE +/- 0.08, N = 8SE +/- 0.07, N = 8SE +/- 0.04, N = 8SE +/- 0.08, N = 8SE +/- 0.09, N = 8SE +/- 0.09, N = 8SE +/- 0.11, N = 7SE +/- 0.17, N = 8SE +/- 0.14, N = 8SE +/- 0.13, N = 9SE +/- 0.20, N = 9SE +/- 0.11, N = 9SE +/- 0.07, N = 10SE +/- 0.08, N = 9288.2284.6289.0290.2301.7322.8270.0293.2294.9337.6354.7355.0363.0319.31. (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 AtomsRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX0.03490.06980.10470.13960.1745SE +/- 0.00029, N = 5SE +/- 0.00031, N = 5SE +/- 0.00016, N = 5SE +/- 0.00019, N = 5SE +/- 0.00058, N = 5SE +/- 0.00078, N = 5SE +/- 0.00054, N = 5SE +/- 0.00020, N = 5SE +/- 0.00049, N = 5SE +/- 0.00014, N = 3SE +/- 0.00023, N = 5SE +/- 0.00031, N = 5SE +/- 0.00010, N = 5SE +/- 0.00130, N = 50.155020.152470.143230.138010.135220.133240.154710.129280.126110.124850.120990.120660.119880.13304

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: OpenCL BLAS - dGEMV-TRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3080RTX 3080 TiTITAN RTX80160240320400SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 33253193253343563753033333333773673721. (CXX) g++ options: -fopenmp -O3 -rdynamic -lOpenCL

Mixbench

OpenBenchmarking.orgWatts, Fewer Is BetterMixbench 2020-06-23GPU Power Consumption MonitorRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX306090120150Min: 9.99 / Avg: 36.3 / Max: 83.91Min: 8.69 / Avg: 32.72 / Max: 78.1Min: 23.92 / Avg: 43.78 / Max: 92.24Min: 15 / Avg: 43.32 / Max: 85.96Min: 9.87 / Avg: 40.09 / Max: 95.14Min: 6.43 / Avg: 47.06 / Max: 104.62Min: 16.66 / Avg: 32.41 / Max: 52.09Min: 12.43 / Avg: 36.65 / Max: 60.96Min: 8.64 / Avg: 34.39 / Max: 59.56Min: 22.16 / Avg: 58.03 / Max: 92.85Min: 17.86 / Avg: 65.41 / Max: 106.54Min: 25.14 / Avg: 79.69 / Max: 134.35Min: 20.23 / Avg: 73.33 / Max: 124.31Min: 8.31 / Avg: 49.38 / Max: 112.81

OpenBenchmarking.orgGIOPS Per Watt, More Is BetterMixbench 2020-06-23Backend: NVIDIA CUDA - Benchmark: IntegerRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX70140210280350215.20244.38210.60236.26278.44313.08181.53230.78291.74174.40213.35195.08225.05311.90

OpenBenchmarking.orgGIOPS, More Is BetterMixbench 2020-06-23Backend: NVIDIA CUDA - Benchmark: IntegerRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX4K8K12K16K20KSE +/- 74.45, N = 15SE +/- 44.96, N = 15SE +/- 0.86, N = 14SE +/- 5.84, N = 14SE +/- 1.25, N = 14SE +/- 340.68, N = 15SE +/- 0.32, N = 14SE +/- 0.45, N = 14SE +/- 0.87, N = 14SE +/- 10.76, N = 15SE +/- 265.32, N = 15SE +/- 306.59, N = 15SE +/- 363.78, N = 15SE +/- 379.98, N = 157811.277995.969219.3410234.2511161.7014731.975883.158458.8410033.7810120.8013956.2115545.6316502.9015399.951. (CXX) g++ options: -lm -lstdc++ -lOpenCL -lrt -O2

OpenBenchmarking.orgWatts, Fewer Is BetterMixbench 2020-06-23GPU Power Consumption MonitorRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX306090120150Min: 10.11 / Avg: 34.85 / Max: 76.06Min: 8.64 / Avg: 33.51 / Max: 78.65Min: 23.73 / Avg: 43 / Max: 75.46Min: 15.7 / Avg: 42.8 / Max: 99.04Min: 9.8 / Avg: 39.42 / Max: 71.47Min: 6.48 / Avg: 46.34 / Max: 131.85Min: 16.56 / Avg: 32.54 / Max: 52.17Min: 13.18 / Avg: 37.71 / Max: 62.35Min: 8.82 / Avg: 35.35 / Max: 58.86Min: 22.32 / Avg: 57.48 / Max: 92.48Min: 20.23 / Avg: 67.66 / Max: 111.37Min: 17.99 / Avg: 80.96 / Max: 134.27Min: 18.23 / Avg: 73.63 / Max: 125.87Min: 10.07 / Avg: 51.54 / Max: 120.68

OpenBenchmarking.orgGFLOPS Per Watt, More Is BetterMixbench 2020-06-23Backend: NVIDIA CUDA - Benchmark: Double PrecisionRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX2468106.3606.7535.9626.7447.9888.8285.4766.7688.5435.3076.2355.7956.7128.255

OpenBenchmarking.orgGFLOPS, More Is BetterMixbench 2020-06-23Backend: NVIDIA CUDA - Benchmark: Double PrecisionRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX110220330440550SE +/- 0.24, N = 14SE +/- 0.26, N = 14SE +/- 0.29, N = 14SE +/- 0.41, N = 14SE +/- 1.19, N = 15SE +/- 11.13, N = 15SE +/- 0.07, N = 14SE +/- 0.53, N = 14SE +/- 0.66, N = 14SE +/- 1.51, N = 14SE +/- 8.42, N = 15SE +/- 8.40, N = 15SE +/- 10.54, N = 15SE +/- 13.41, N = 15221.65226.32256.33288.66314.85409.05178.20255.22302.02305.04421.88469.16494.24425.451. (CXX) g++ options: -lm -lstdc++ -lOpenCL -lrt -O2

OpenBenchmarking.orgWatts, Fewer Is BetterMixbench 2020-06-23GPU Power Consumption MonitorRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX306090120150Min: 10.09 / Avg: 35.73 / Max: 82.38Min: 9.37 / Avg: 32.3 / Max: 78.95Min: 24.07 / Avg: 44.35 / Max: 95.72Min: 16.5 / Avg: 44.34 / Max: 105.67Min: 10.04 / Avg: 36.71 / Max: 89.26Min: 7.7 / Avg: 49.34 / Max: 121.15Min: 16.88 / Avg: 32.8 / Max: 52.19Min: 13.92 / Avg: 37.89 / Max: 62.71Min: 9.03 / Avg: 35.52 / Max: 60.82Min: 22.44 / Avg: 59 / Max: 95.96Min: 26.73 / Avg: 69.58 / Max: 113.82Min: 15.39 / Avg: 81.94 / Max: 138.88Min: 23.78 / Avg: 75.64 / Max: 131.24Min: 11.54 / Avg: 52.89 / Max: 124.22

OpenBenchmarking.orgGFLOPS Per Watt, More Is BetterMixbench 2020-06-23Backend: NVIDIA CUDA - Benchmark: Single PrecisionRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX130260390520650230.25259.33217.99243.67316.62301.00390.76481.61602.37361.63419.95398.39451.72291.75

OpenBenchmarking.orgGFLOPS, More Is BetterMixbench 2020-06-23Backend: NVIDIA CUDA - Benchmark: Single PrecisionRTX 2060 SUPERRTX 2070RTX 2070 SUPERRTX 2080RTX 2080 SUPERRTX 2080 TiRTX 3060RTX 3060 TiRTX 3070RTX 3070 TiRTX 3080RTX 3080 TiRTX 3090TITAN RTX7K14K21K28K35KSE +/- 12.37, N = 14SE +/- 6.70, N = 14SE +/- 5.79, N = 14SE +/- 12.22, N = 15SE +/- 124.12, N = 15SE +/- 497.84, N = 15SE +/- 1.31, N = 14SE +/- 9.34, N = 14SE +/- 177.14, N = 15SE +/- 213.66, N = 15SE +/- 546.33, N = 15SE +/- 563.16, N = 15SE +/- 561.56, N = 15SE +/- 520.86, N = 158225.598377.739668.0510803.6311622.1814851.9112815.8318247.6121395.3721337.8329218.8532644.6834169.7415431.561. (CXX) g++ options: -lm -lstdc++ -lOpenCL -lrt -O2

84 Results Shown

Blender:
  Classroom - CUDA
  Barbershop - NVIDIA OptiX
clpeak
Blender:
  Classroom - NVIDIA OptiX
  Pabellon Barcelona - CUDA
  Pabellon Barcelona - NVIDIA OptiX
Chaos Group V-RAY:
  NVIDIA CUDA GPU
  NVIDIA RTX GPU
LuxCoreRender
Blender:
  Fishy Cat - CUDA
  BMW27 - CUDA
LuxCoreRender
vkpeak
ArrayFire
Blender:
  BMW27 - NVIDIA OptiX
  Fishy Cat - NVIDIA OptiX
LuxCoreRender
vkpeak:
  fp64-vec4
  fp32-scalar
  fp64-scalar
clpeak
vkpeak:
  int32-scalar
  int16-scalar
  fp16-scalar
  int32-vec4
  fp16-vec4
SHOC Scalable HeterOgeneous Computing
Blender
ViennaCL
Hashcat
ViennaCL
Hashcat
SHOC Scalable HeterOgeneous Computing
Hashcat
clpeak
vkpeak
Hashcat
IndigoBench
LuxCoreRender
Mixbench
OctaneBench
LuxCoreRender
LeelaChessZero
RedShift Demo
RealSR-NCNN
Hashcat
PlaidML:
  No - Inference - VGG19 - OpenCL
  No - Inference - VGG16 - OpenCL
cl-mem
clpeak
SHOC Scalable HeterOgeneous Computing:
  OpenCL - FFT SP
  OpenCL - S3D
VkResample
ViennaCL
cl-mem
PlaidML
IndigoBench
ViennaCL
Betsy GPU Compressor
RealSR-NCNN
ViennaCL
SHOC Scalable HeterOgeneous Computing
Betsy GPU Compressor
ViennaCL:
  OpenCL BLAS - dDOT
  OpenCL BLAS - dCOPY
FAHBench
ViennaCL
ArrayFire
Waifu2x-NCNN Vulkan
ViennaCL:
  OpenCL BLAS - sAXPY
  OpenCL BLAS - sDOT
  OpenCL BLAS - sCOPY
cl-mem
NAMD CUDA
ViennaCL
Mixbench:
  GPU Power Consumption Monitor
  NVIDIA CUDA - Integer
Mixbench
Mixbench:
  GPU Power Consumption Monitor
  NVIDIA CUDA - Double Precision
Mixbench
Mixbench:
  GPU Power Consumption Monitor
  NVIDIA CUDA - Single Precision
Mixbench