GeForce RTX 3090 Windows vs. Ubuntu Linux

GeForce RTX 3090 benchmarks on Windows and Linux for a future article on Phoronix.com by Michael Larabel.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2105048-IB-RTXWINDOW53
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

Limit displaying results to tests within:

CPU Massive 2 Tests
Creator Workloads 4 Tests
Multi-Core 4 Tests
NVIDIA GPU Compute 11 Tests
Renderers 4 Tests
Vulkan Compute 4 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

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
Windows 10
May 02 2021
  3 Hours, 58 Minutes
Ubuntu 21.04
May 04 2021
  3 Hours, 55 Minutes
Invert Hiding All Results Option
  3 Hours, 57 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):


GeForce RTX 3090 Windows vs. Ubuntu LinuxProcessorMotherboardMemoryDiskGraphicsNetworkChipsetAudioMonitorOSKernelDisplay DriverOpenCLFile-SystemScreen ResolutionDesktopDisplay ServerOpenGLVulkanCompilerWindows 10Ubuntu 21.04AMD Ryzen 9 5900X 12-Core @ 3.70GHz (12 Cores / 24 Threads)ASUS ROG CROSSHAIR VIII HERO (3402 BIOS)2 x 8192 MB 3600MHz F4-3600C16-8GTRG932GB Sabrent Rocket 4.0 Plus + 15GB Corsair Voyager 3.0 USBNVIDIA GeForce RTX 3090 24GBIntel I211 Gigabit ConnectionMicrosoft Windows 10 Pro Build 1904310.0 (x86_64)466.27 (27.21.14.6627)OpenCL 3.0 CUDA 11.3.101NTFS3840x2160AMD Starship/Matisse16GB1000GB Sabrent Rocket 4.0 PlusNVIDIA GA102 HD AudioASUS VP28URealtek RTL8125 2.5GbE + Intel I211Ubuntu 21.045.11.0-16-generic (x86_64)GNOME Shell 3.38.4X Server 1.20.11NVIDIA 465.274.6.01.2.168GCC 10.3.0ext4OpenBenchmarking.orgSecurity Details- Windows 10: __user pointer sanitization: Disabled + Retpoline: Full + IBPB: Always + IBRS: Enabled + STIBP: Enabled- Ubuntu 21.04: 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 Kernel Details- Ubuntu 21.04: Transparent Huge Pages: madviseCompiler Details- Ubuntu 21.04: --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 Details- Ubuntu 21.04: Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0xa201009OpenCL Details- Ubuntu 21.04: GPU Compute Cores: 10496

Windows 10 vs. Ubuntu 21.04 ComparisonPhoronix Test SuiteBaseline+10.3%+10.3%+20.6%+20.6%+30.9%+30.9%41.1%24.2%12.3%8.5%6.5%5.6%4%3.5%3.4%3%2.2%NVIDIA CUDA GPULuxCore Benchmark - GPU21.2%Danish Mood - GPU20.4%DLSC - GPU16%4x - No13.8%R.C.a.P - GPUBarbershop - NVIDIA OptiX7%Fishy Cat - CUDABMW27 - NVIDIA OptiXOrange Juice - GPUClassroom - NVIDIA OptiX3.9%Fishy Cat - NVIDIA OptiXBarbershop - CUDABMW27 - CUDAPabellon Barcelona - NVIDIA OptiXFAHBenchChaos Group V-RAYLuxCoreRenderLuxCoreRenderLuxCoreRenderRealSR-NCNNRedShift DemoLuxCoreRenderBlenderBlenderBlenderLuxCoreRenderBlenderBlenderBlenderBlenderBlenderWindows 10Ubuntu 21.04

GeForce RTX 3090 Windows vs. Ubuntu Linuxblender: BMW27 - CUDAblender: Classroom - CUDAblender: Fishy Cat - CUDAblender: Barbershop - CUDAblender: BMW27 - NVIDIA OptiXblender: Classroom - NVIDIA OptiXblender: Fishy Cat - NVIDIA OptiXblender: Barbershop - NVIDIA OptiXblender: Pabellon Barcelona - CUDAblender: Pabellon Barcelona - NVIDIA OptiXv-ray: NVIDIA RTX GPUv-ray: NVIDIA CUDA GPUfahbench: indigobench: OpenCL GPU - Bedroomindigobench: OpenCL GPU - Supercarluxcorerender: DLSC - GPUluxcorerender: Danish Mood - GPUluxcorerender: Orange Juice - GPUluxcorerender: LuxCore Benchmark - GPUluxcorerender: Rainbow Colors and Prism - GPUoctanebench: Total Scorerealsr-ncnn: 4x - Norealsr-ncnn: 4x - Yesredshift: vkpeak: fp32-scalarvkpeak: fp32-vec4vkpeak: fp16-scalarvkpeak: fp16-vec4vkpeak: fp64-scalarvkpeak: fp64-vec4vkpeak: int32-scalarvkpeak: int32-vec4vkpeak: int16-scalarvkpeak: int16-vec4vkresample: 2x - Singlewaifu2x-ncnn: 2x - 3 - YesWindows 10Ubuntu 21.0418.8851.7236.91387.2012.9929.4622.64323.77120.8045.3121641688243.735220.74552.41413.3410.619.8413.2429.33676.6146414.89228.04715520868.2727560.6820694.2641706.09650.93652.2720638.7720572.9913603.6716881.999.3143.27618.3351.4534.65374.6012.3030.6021.87346.39120.3144.3221652096343.798620.75152.73711.508.8110.2310.9231.82682.4549595.56828.34613820908.2627526.7920791.2141717.51651.98653.4820764.1020642.0213673.6216872.759.2683.286OpenBenchmarking.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: BMW27 - Compute: CUDAWindows 10Ubuntu 21.04510152025SE +/- 0.01, N = 3SE +/- 0.04, N = 318.8818.33
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: BMW27 - Compute: CUDAWindows 10Ubuntu 21.04510152025Min: 18.87 / Avg: 18.88 / Max: 18.89Min: 18.29 / Avg: 18.33 / Max: 18.4

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Classroom - Compute: CUDAWindows 10Ubuntu 21.041224364860SE +/- 0.03, N = 3SE +/- 0.04, N = 351.7251.45
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Classroom - Compute: CUDAWindows 10Ubuntu 21.041020304050Min: 51.65 / Avg: 51.72 / Max: 51.75Min: 51.37 / Avg: 51.45 / Max: 51.49

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Fishy Cat - Compute: CUDAWindows 10Ubuntu 21.04816243240SE +/- 0.01, N = 3SE +/- 0.01, N = 336.9134.65
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Fishy Cat - Compute: CUDAWindows 10Ubuntu 21.04816243240Min: 36.9 / Avg: 36.91 / Max: 36.92Min: 34.63 / Avg: 34.65 / Max: 34.66

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Barbershop - Compute: CUDAWindows 10Ubuntu 21.0480160240320400SE +/- 0.28, N = 3SE +/- 0.06, N = 3387.20374.60
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Barbershop - Compute: CUDAWindows 10Ubuntu 21.0470140210280350Min: 386.78 / Avg: 387.2 / Max: 387.73Min: 374.53 / Avg: 374.6 / Max: 374.72

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: BMW27 - Compute: NVIDIA OptiXWindows 10Ubuntu 21.043691215SE +/- 3.10, N = 15SE +/- 2.77, N = 1512.9912.30
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: BMW27 - Compute: NVIDIA OptiXWindows 10Ubuntu 21.0448121620Min: 9.86 / Avg: 12.99 / Max: 56.42Min: 9.49 / Avg: 12.3 / Max: 51.07

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Classroom - Compute: NVIDIA OptiXWindows 10Ubuntu 21.04714212835SE +/- 0.11, N = 3SE +/- 0.12, N = 329.4630.60
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Classroom - Compute: NVIDIA OptiXWindows 10Ubuntu 21.04714212835Min: 29.24 / Avg: 29.46 / Max: 29.58Min: 30.36 / Avg: 30.6 / Max: 30.74

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Fishy Cat - Compute: NVIDIA OptiXWindows 10Ubuntu 21.04510152025SE +/- 3.08, N = 15SE +/- 2.70, N = 1522.6421.87
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Fishy Cat - Compute: NVIDIA OptiXWindows 10Ubuntu 21.04510152025Min: 19.49 / Avg: 22.64 / Max: 65.69Min: 19.11 / Avg: 21.87 / Max: 59.65

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Barbershop - Compute: NVIDIA OptiXWindows 10Ubuntu 21.0480160240320400SE +/- 0.48, N = 3SE +/- 0.13, N = 3323.77346.39
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Barbershop - Compute: NVIDIA OptiXWindows 10Ubuntu 21.0460120180240300Min: 322.98 / Avg: 323.77 / Max: 324.65Min: 346.22 / Avg: 346.39 / Max: 346.65

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Pabellon Barcelona - Compute: CUDAWindows 10Ubuntu 21.04306090120150SE +/- 0.05, N = 3SE +/- 0.07, N = 3120.80120.31
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Pabellon Barcelona - Compute: CUDAWindows 10Ubuntu 21.0420406080100Min: 120.71 / Avg: 120.8 / Max: 120.87Min: 120.17 / Avg: 120.31 / Max: 120.39

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Pabellon Barcelona - Compute: NVIDIA OptiXWindows 10Ubuntu 21.041020304050SE +/- 0.00, N = 3SE +/- 0.12, N = 345.3144.32
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.92Blend File: Pabellon Barcelona - Compute: NVIDIA OptiXWindows 10Ubuntu 21.04918273645Min: 45.3 / Avg: 45.31 / Max: 45.31Min: 44.15 / Avg: 44.32 / Max: 44.55

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 GPUWindows 10Ubuntu 21.045001000150020002500SE +/- 3.48, N = 3SE +/- 1.15, N = 321642165
OpenBenchmarking.orgvrays, More Is BetterChaos Group V-RAY 5Mode: NVIDIA RTX GPUWindows 10Ubuntu 21.04400800120016002000Min: 2158 / Avg: 2163.67 / Max: 2170Min: 2163 / Avg: 2165 / Max: 2167

OpenBenchmarking.orgvpaths, More Is BetterChaos Group V-RAY 5Mode: NVIDIA CUDA GPUWindows 10Ubuntu 21.04400800120016002000SE +/- 2.33, N = 3SE +/- 1.00, N = 316882096
OpenBenchmarking.orgvpaths, More Is BetterChaos Group V-RAY 5Mode: NVIDIA CUDA GPUWindows 10Ubuntu 21.04400800120016002000Min: 1683 / Avg: 1687.67 / Max: 1690Min: 2095 / Avg: 2096 / Max: 2098

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 2.3.2Windows 10Ubuntu 21.0470140210280350SE +/- 0.12, N = 3SE +/- 0.20, N = 3243.74343.80
OpenBenchmarking.orgNs Per Day, More Is BetterFAHBench 2.3.2Windows 10Ubuntu 21.0460120180240300Min: 243.52 / Avg: 243.74 / Max: 243.92Min: 343.44 / Avg: 343.8 / Max: 344.13

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: BedroomWindows 10Ubuntu 21.04510152025SE +/- 0.01, N = 3SE +/- 0.02, N = 320.7520.75
OpenBenchmarking.orgM samples/s, More Is BetterIndigoBench 4.4Acceleration: OpenCL GPU - Scene: BedroomWindows 10Ubuntu 21.04510152025Min: 20.74 / Avg: 20.75 / Max: 20.76Min: 20.73 / Avg: 20.75 / Max: 20.8

OpenBenchmarking.orgM samples/s, More Is BetterIndigoBench 4.4Acceleration: OpenCL GPU - Scene: SupercarWindows 10Ubuntu 21.041224364860SE +/- 0.03, N = 3SE +/- 0.10, N = 352.4152.74
OpenBenchmarking.orgM samples/s, More Is BetterIndigoBench 4.4Acceleration: OpenCL GPU - Scene: SupercarWindows 10Ubuntu 21.041122334455Min: 52.37 / Avg: 52.41 / Max: 52.46Min: 52.55 / Avg: 52.74 / Max: 52.88

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: GPUWindows 10Ubuntu 21.043691215SE +/- 0.03, N = 3SE +/- 0.03, N = 313.3411.50MIN: 13.18 / MAX: 13.5MIN: 10.31 / MAX: 11.69
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: DLSC - Acceleration: GPUWindows 10Ubuntu 21.0448121620Min: 13.31 / Avg: 13.34 / Max: 13.41Min: 11.45 / Avg: 11.5 / Max: 11.56

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Danish Mood - Acceleration: GPUWindows 10Ubuntu 21.043691215SE +/- 0.04, N = 3SE +/- 0.08, N = 310.618.81MIN: 10.06 / MAX: 10.98MIN: 2.72 / MAX: 10.68
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Danish Mood - Acceleration: GPUWindows 10Ubuntu 21.043691215Min: 10.54 / Avg: 10.61 / Max: 10.67Min: 8.69 / Avg: 8.81 / Max: 8.96

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Orange Juice - Acceleration: GPUWindows 10Ubuntu 21.043691215SE +/- 0.07, N = 3SE +/- 0.02, N = 39.8410.23MIN: 9.41 / MAX: 10.31MIN: 8.32 / MAX: 13.46
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Orange Juice - Acceleration: GPUWindows 10Ubuntu 21.043691215Min: 9.73 / Avg: 9.84 / Max: 9.98Min: 10.2 / Avg: 10.23 / Max: 10.28

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: LuxCore Benchmark - Acceleration: GPUWindows 10Ubuntu 21.043691215SE +/- 0.04, N = 3SE +/- 0.02, N = 313.2410.92MIN: 12.65 / MAX: 13.62MIN: 3.32 / MAX: 13
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: LuxCore Benchmark - Acceleration: GPUWindows 10Ubuntu 21.0448121620Min: 13.17 / Avg: 13.24 / Max: 13.31Min: 10.9 / Avg: 10.92 / Max: 10.95

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Rainbow Colors and Prism - Acceleration: GPUWindows 10Ubuntu 21.04714212835SE +/- 0.40, N = 15SE +/- 0.24, N = 1229.3331.82MIN: 16.77 / MAX: 35.64MIN: 22.56 / MAX: 34.85
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Rainbow Colors and Prism - Acceleration: GPUWindows 10Ubuntu 21.04714212835Min: 26.89 / Avg: 29.33 / Max: 31.57Min: 29.28 / Avg: 31.82 / Max: 32.39

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 ScoreWindows 10Ubuntu 21.04150300450600750676.61682.45

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: NoWindows 10Ubuntu 21.041.25282.50563.75845.01126.264SE +/- 0.015, N = 3SE +/- 0.010, N = 34.8925.568
OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: NoWindows 10Ubuntu 21.04246810Min: 4.87 / Avg: 4.89 / Max: 4.92Min: 5.55 / Avg: 5.57 / Max: 5.59

OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: YesWindows 10Ubuntu 21.04714212835SE +/- 0.05, N = 3SE +/- 0.04, N = 328.0528.35
OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: YesWindows 10Ubuntu 21.04612182430Min: 27.95 / Avg: 28.05 / Max: 28.13Min: 28.3 / Avg: 28.35 / Max: 28.42

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 3.0Windows 10Ubuntu 21.04306090120150155138

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-scalarWindows 10Ubuntu 21.044K8K12K16K20KSE +/- 108.41, N = 3SE +/- 118.56, N = 320868.2720908.26
OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp32-scalarWindows 10Ubuntu 21.044K8K12K16K20KMin: 20753.26 / Avg: 20868.27 / Max: 21084.94Min: 20784.85 / Avg: 20908.26 / Max: 21145.3

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp32-vec4Windows 10Ubuntu 21.046K12K18K24K30KSE +/- 117.24, N = 3SE +/- 137.30, N = 327560.6827526.79
OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp32-vec4Windows 10Ubuntu 21.045K10K15K20K25KMin: 27356.57 / Avg: 27560.68 / Max: 27762.67Min: 27389.45 / Avg: 27526.79 / Max: 27801.4

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp16-scalarWindows 10Ubuntu 21.044K8K12K16K20KSE +/- 109.48, N = 3SE +/- 89.50, N = 320694.2620791.21
OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp16-scalarWindows 10Ubuntu 21.044K8K12K16K20KMin: 20584.38 / Avg: 20694.26 / Max: 20913.22Min: 20640.1 / Avg: 20791.21 / Max: 20949.87

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp16-vec4Windows 10Ubuntu 21.049K18K27K36K45KSE +/- 238.97, N = 3SE +/- 206.78, N = 341706.0941717.51
OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp16-vec4Windows 10Ubuntu 21.047K14K21K28K35KMin: 41465.59 / Avg: 41706.09 / Max: 42184.02Min: 41509.41 / Avg: 41717.51 / Max: 42131.07

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp64-scalarWindows 10Ubuntu 21.04140280420560700SE +/- 2.87, N = 3SE +/- 3.25, N = 3650.93651.98
OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp64-scalarWindows 10Ubuntu 21.04110220330440550Min: 648.06 / Avg: 650.93 / Max: 656.68Min: 648.71 / Avg: 651.98 / Max: 658.48

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp64-vec4Windows 10Ubuntu 21.04140280420560700SE +/- 4.19, N = 2SE +/- 2.77, N = 3652.27653.48
OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20210424fp64-vec4Windows 10Ubuntu 21.04120240360480600Min: 648.07 / Avg: 652.27 / Max: 656.46Min: 648.72 / Avg: 653.48 / Max: 658.31

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int32-scalarWindows 10Ubuntu 21.044K8K12K16K20KSE +/- 85.18, N = 3SE +/- 89.31, N = 320638.7720764.10
OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int32-scalarWindows 10Ubuntu 21.044K8K12K16K20KMin: 20553.42 / Avg: 20638.77 / Max: 20809.12Min: 20613.19 / Avg: 20764.1 / Max: 20922.31

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int32-vec4Windows 10Ubuntu 21.044K8K12K16K20KSE +/- 99.15, N = 3SE +/- 64.85, N = 320572.9920642.02
OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int32-vec4Windows 10Ubuntu 21.044K8K12K16K20KMin: 20473.79 / Avg: 20572.99 / Max: 20771.28Min: 20518.52 / Avg: 20642.02 / Max: 20738.06

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int16-scalarWindows 10Ubuntu 21.043K6K9K12K15KSE +/- 34.81, N = 3SE +/- 32.44, N = 313603.6713673.62
OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int16-scalarWindows 10Ubuntu 21.042K4K6K8K10KMin: 13568.65 / Avg: 13603.67 / Max: 13673.28Min: 13608.76 / Avg: 13673.62 / Max: 13707.36

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int16-vec4Windows 10Ubuntu 21.044K8K12K16K20KSE +/- 43.37, N = 3SE +/- 73.46, N = 316881.9916872.75
OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20210424int16-vec4Windows 10Ubuntu 21.043K6K9K12K15KMin: 16838.33 / Avg: 16881.99 / Max: 16968.72Min: 16745.34 / Avg: 16872.75 / Max: 16999.82

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: SingleWindows 10Ubuntu 21.043691215SE +/- 0.009, N = 3SE +/- 0.006, N = 39.3149.2681. (CXX) g++ options: -O3 -pthread
OpenBenchmarking.orgms, Fewer Is BetterVkResample 1.0Upscale: 2x - Precision: SingleWindows 10Ubuntu 21.043691215Min: 9.3 / Avg: 9.31 / Max: 9.32Min: 9.26 / Avg: 9.27 / Max: 9.281. (CXX) g++ options: -O3 -pthread

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: YesWindows 10Ubuntu 21.040.73941.47882.21822.95763.697SE +/- 0.046, N = 15SE +/- 0.003, N = 33.2763.286
OpenBenchmarking.orgSeconds, Fewer Is BetterWaifu2x-NCNN Vulkan 20200818Scale: 2x - Denoise: 3 - TAA: YesWindows 10Ubuntu 21.04246810Min: 3.2 / Avg: 3.28 / Max: 3.92Min: 3.28 / Avg: 3.29 / Max: 3.29