vulkan-benchmarks

AMD Ryzen 9 7950X 16-Core testing with a ASUS ROG STRIX X670E-E GAMING WIFI (1416 BIOS) and NVIDIA GeForce RTX 4090 24GB on Ubuntu 23.04 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2308069-PTS-VULKANBE16
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
Allow Limiting Results To Certain Suite(s)

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
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
Toggle/Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
a
August 01 2023
  3 Hours, 11 Minutes
b
August 01 2023
  1 Hour, 30 Minutes
c
August 01 2023
  1 Hour, 32 Minutes
d
August 01 2023
  3 Hours, 45 Minutes
e
August 01 2023
  3 Hours, 16 Minutes
f
August 02 2023
  1 Hour, 53 Minutes
g
August 02 2023
  2 Hours, 9 Minutes
h
August 02 2023
  47 Minutes
i
August 02 2023
  1 Hour, 50 Minutes
4080
August 02 2023
  2 Hours, 4 Minutes
4080 rep
August 02 2023
  2 Hours, 7 Minutes
4080 xxx
August 02 2023
  2 Hours, 8 Minutes
4080 zzz
August 02 2023
  2 Hours, 9 Minutes
3090
August 03 2023
  2 Hours, 44 Minutes
3090 rep
August 03 2023
  2 Hours, 54 Minutes
3070
August 03 2023
  4 Hours, 56 Minutes
RTX 3070 Ti
August 04 2023
  1 Day, 7 Hours, 27 Minutes
4090
August 06 2023
  2 Hours, 52 Minutes
4090 rep
August 06 2023
  2 Hours, 54 Minutes
nv 4090
August 06 2023
  2 Hours, 52 Minutes
Invert Behavior (Only Show Selected Data)
  3 Hours, 57 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):


vulkan-benchmarks ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerOpenGLCompilerFile-SystemScreen ResolutionDisplay Driverabcdefghi40804080 rep4080 xxx4080 zzz30903090 rep3070RTX 3070 Ti40904090 repnv 4090AMD Ryzen 9 7950X 16-Core @ 4.50GHz (16 Cores / 32 Threads)ASUS ROG STRIX X670E-E GAMING WIFI (1416 BIOS)AMD Device 14d832GBWestern Digital WD_BLACK SN850X 1000GB + 4001GBAMD Radeon RX 6700 XT (2855/1000MHz)AMD Navi 21/23ASUS MG28UIntel I225-V + Intel Wi-Fi 6 AX210/AX211/AX411Ubuntu 23.046.4.6-060406-generic (x86_64)GNOME Shell 44.2X Server 1.21.1.7 + Wayland4.6 Mesa 23.3~git2307260600.87109c~oibaf~l (git-87109c3 2023-07-26 lunar-oibaf-ppa) (LLVM 15.0.7 DRM 3.52)GCC 12.2.0ext43840x2160MSI NVIDIA GeForce RTX 4060 8GBNVIDIA Device 22beX Server 1.21.1.7NVIDIA 535.86.054.6.0eVGA NVIDIA GeForce RTX 3060 12GBNVIDIA GA106 HD AudioNVIDIA GeForce RTX 3060 Ti 8GBNVIDIA GA104 HD Audio2560x1440NVIDIA GeForce RTX 4080 16GBNVIDIA Device 22bb3840x2160NVIDIA GeForce RTX 3090 24GBNVIDIA GA102 HD AudioNVIDIA GeForce RTX 3070 8GBNVIDIA GA104 HD Audio2560x1440NVIDIA GeForce RTX 3070 Ti 8GBNVIDIA GeForce RTX 4090 24GBNVIDIA AD102 HD Audio3840x2160OpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseCompiler Details- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-Pa930Z/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-Pa930Z/gcc-12-12.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Processor Details- a: Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0xa601203- b: Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0xa601203- c: Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0xa601203- d: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- e: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- f: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- g: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- h: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- i: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- 4080: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- 4080 rep: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- 4080 xxx: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- 4080 zzz: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- 3090: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- 3090 rep: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- 3070: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- RTX 3070 Ti: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- 4090: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- 4090 rep: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203- nv 4090: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa601203Graphics Details- a: BAR1 / Visible vRAM Size: 12272 MB - vBIOS Version: 113-D5121100-101- b: BAR1 / Visible vRAM Size: 12272 MB - vBIOS Version: 113-D5121100-101- c: BAR1 / Visible vRAM Size: 12272 MB - vBIOS Version: 113-D5121100-101- d: BAR1 / Visible vRAM Size: 8192 MiB - vBIOS Version: 95.07.31.00.e3- e: BAR1 / Visible vRAM Size: 8192 MiB - vBIOS Version: 95.07.31.00.e3- f: BAR1 / Visible vRAM Size: 16384 MiB - vBIOS Version: 94.06.14.40.46- g: BAR1 / Visible vRAM Size: 16384 MiB - vBIOS Version: 94.06.14.40.46- h: BAR1 / Visible vRAM Size: 16384 MiB - vBIOS Version: 94.06.14.40.46- i: BAR1 / Visible vRAM Size: 8192 MiB - vBIOS Version: 94.04.25.00.2c- 4080: BAR1 / Visible vRAM Size: 16384 MiB - vBIOS Version: 95.03.0e.00.04- 4080 rep: BAR1 / Visible vRAM Size: 16384 MiB - vBIOS Version: 95.03.0e.00.04- 4080 xxx: BAR1 / Visible vRAM Size: 16384 MiB - vBIOS Version: 95.03.0e.00.04- 4080 zzz: BAR1 / Visible vRAM Size: 16384 MiB - vBIOS Version: 95.03.0e.00.04- 3090: BAR1 / Visible vRAM Size: 32768 MiB - vBIOS Version: 94.02.27.00.02- 3090 rep: BAR1 / Visible vRAM Size: 32768 MiB - vBIOS Version: 94.02.27.00.02- 3070: BAR1 / Visible vRAM Size: 8192 MiB - vBIOS Version: 94.04.25.00.2b- RTX 3070 Ti: BAR1 / Visible vRAM Size: 8192 MiB - vBIOS Version: 94.04.5b.00.02- 4090: BAR1 / Visible vRAM Size: 32768 MiB - vBIOS Version: 95.02.20.00.01- 4090 rep: BAR1 / Visible vRAM Size: 32768 MiB - vBIOS Version: 95.02.20.00.01- nv 4090: BAR1 / Visible vRAM Size: 32768 MiB - vBIOS Version: 95.02.20.00.01Security Details- itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

vulkan-benchmarks vkpeak: int32-scalarvkpeak: int32-vec4vkfft: FFT + iFFT C2C 1D batched in double precisionvkfft: FFT + iFFT C2C Bluestein benchmark in double precisionvkpeak: fp64-vec4vkpeak: fp64-scalarvkpeak: int16-vec4vkresample: 2x - Singlevkfft: FFT + iFFT C2C 1D batched in single precisionvkfft: FFT + iFFT C2C 1D batched in single precision, no reshufflingvkfft: FFT + iFFT C2C 1D batched in half precisionncnn: CPU - FastestDetvkfft: FFT + iFFT R2C / C2Rvkfft: FFT + iFFT C2C multidimensional in single precisionncnn: CPU-v3-v3-v3 - FastestDetvkpeak: fp32-scalarvkpeak: fp32-vec4ncnn: Vulkan GPU-v3-v3-v3-v3-v3 - mobilenet-v3vkpeak: fp16-scalarncnn: CPU-v2-v2 - mobilenet-v2vkpeak: fp16-vec4vkpeak: int16-scalarvkresample: 2x - Doublencnn: Vulkan GPU-v3-v3-v3 - blazefacencnn: CPU-v3-v3-v3-v2-v2 - mobilenet-v2ncnn: Vulkan GPU - blazefacencnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mnasnetncnn: CPU - mobilenetncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - alexnetncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mnasnetvkfft: FFT + iFFT C2C Bluestein in single precisionncnn: CPU - resnet18ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - squeezenet_ssdncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - alexnetncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v2-v2 - mobilenet-v2ncnn: Vulkan GPU-v3-v3-v3 - googlenetncnn: Vulkan GPU - mobilenetncnn: CPU-v3-v3-v3 - blazefacencnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - blazefacencnn: CPU-v3-v3-v3 - vision_transformerncnn: CPU-v3-v3-v3 - alexnetncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - resnet18ncnn: Vulkan GPU-v3-v3-v3 - resnet18ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - efficientnet-b0ncnn: CPU-v3-v3-v3 - squeezenet_ssdncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - FastestDetncnn: Vulkan GPU - alexnetncnn: CPU - mnasnetncnn: Vulkan GPU - googlenetncnn: CPU-v3-v3-v3-v3-v3-v3 - FastestDetncnn: Vulkan GPU-v2-v2 - mobilenet-v2ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - alexnetncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - efficientnet-b0ncnn: CPU-v3-v3-v3 - regnety_400mncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - FastestDetncnn: CPU-v3-v3-v3-v3-v3-v3 - alexnetncnn: CPU-v3-v3-v3 - shufflenet-v2ncnn: Vulkan GPU - resnet18ncnn: CPU-v3-v3-v3-v3-v3-v3 - resnet18ncnn: CPU-v3-v3-v3-v3-v3-v3 - googlenetncnn: CPU - vgg16ncnn: CPU-v3-v3-v3 - resnet50ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - shufflenet-v2ncnn: CPU - efficientnet-b0ncnn: CPU-v3-v3 - mobilenet-v3ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - googlenetncnn: CPU-v3-v3-v3-v3-v3-v3 - efficientnet-b0ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - googlenetncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - blazefacencnn: CPU - vision_transformerncnn: Vulkan GPU-v3-v3-v3 - mnasnetncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - googlenetncnn: Vulkan GPU-v3-v3-v3 - vgg16ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - vgg16ncnn: CPU-v3-v3-v3 - efficientnet-b0ncnn: Vulkan GPU - efficientnet-b0ncnn: CPU-v3-v3-v3-v3-v3-v3 - shufflenet-v2ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - resnet50ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - resnet50ncnn: Vulkan GPU - resnet50ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - resnet50ncnn: CPU - blazefacencnn: Vulkan GPU-v3-v3-v3 - resnet50ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - resnet18ncnn: Vulkan GPU-v3-v3-v3 - FastestDetncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - alexnetncnn: CPU - googlenetncnn: CPU-v3-v3-v3 - mnasnetncnn: Vulkan GPU-v3-v3-v3 - vision_transformerncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - yolov4-tinyncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenetncnn: Vulkan GPU-v3-v3-v3 - yolov4-tinyncnn: Vulkan GPU-v3-v3-v3-v2-v2 - mobilenet-v2ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - resnet18ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenet-v3ncnn: Vulkan GPU-v3-v3-v3 - squeezenet_ssdncnn: Vulkan GPU-v3-v3-v3 - alexnetncnn: CPU-v3-v3-v3 - yolov4-tinyncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - mobilenetncnn: CPU-v3-v3-v3-v3-v3-v3-v2-v2 - mobilenet-v2ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - yolov4-tinyncnn: CPU-v3-v3-v3-v3-v3-v3 - regnety_400mncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - vgg16ncnn: Vulkan GPU - squeezenet_ssdncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - regnety_400mncnn: CPU - alexnetncnn: Vulkan GPU - yolov4-tinyncnn: Vulkan GPU-v3-v3-v3 - regnety_400mncnn: CPU - regnety_400mncnn: CPU-v3-v3-v3 - mobilenetncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - yolov4-tinyncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - regnety_400mncnn: Vulkan GPU - vision_transformerncnn: CPU-v3-v3-v3-v3-v3-v3 - yolov4-tinyncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - vision_transformerncnn: CPU-v3-v3-v3-v3-v3-v3 - resnet50ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - vision_transformerncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - vision_transformerncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - efficientnet-b0ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - squeezenet_ssdncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - squeezenet_ssdncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - vgg16ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - regnety_400mncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - resnet18ncnn: CPU-v3-v3-v3 - googlenetncnn: CPU - yolov4-tinyncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - squeezenet_ssdncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - googlenetncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - vision_transformerncnn: CPU - resnet50ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - resnet50ncnn: CPU-v3-v3-v3-v3-v3-v3 - vgg16ncnn: CPU-v3-v3-v3 - resnet18ncnn: Vulkan GPU - FastestDetncnn: Vulkan GPU - shufflenet-v2ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenetncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - vgg16ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenetncnn: CPU-v3-v3-v3 - vgg16ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - regnety_400mncnn: Vulkan GPU-v3-v3 - mobilenet-v3ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenet-v3ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenet-v3ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mnasnetncnn: Vulkan GPU - vgg16ncnn: Vulkan GPU-v3-v3-v3 - mobilenetncnn: CPU-v3-v3-v3-v3-v3-v3 - vision_transformerncnn: CPU-v3-v3-v3-v3-v3-v3 - mnasnetncnn: CPU-v3-v3-v3-v3-v3-v3 - mobilenetncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - yolov4-tinyncnn: Vulkan GPU-v3-v3-v3 - efficientnet-b0ncnn: CPU - shufflenet-v2ncnn: Vulkan GPU - regnety_400mncnn: CPU-v3-v3-v3-v3-v3-v3 - squeezenet_ssdncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - blazefacencnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenet-v3ncnn: CPU-v3-v3-v3-v3-v3-v3 - blazefacencnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenet-v3ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v2-v2 - mobilenet-v2ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - shufflenet-v2ncnn: CPU - squeezenet_ssdncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - FastestDetncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - shufflenet-v2ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v2-v2 - mobilenet-v2ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v2-v2 - mobilenet-v2ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - efficientnet-b0ncnn: Vulkan GPU - mnasnetncnn: CPU-v3-v3-v3-v3-v3 - mobilenet-v3ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - shufflenet-v2ncnn: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - FastestDetncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - mnasnetncnn: Vulkan GPU-v3-v3-v3 - shufflenet-v2ncnn: Vulkan GPU-v3-v3-v3-v3-v3-v3 - blazefaceabcdefghi40804080 rep4080 xxx4080 zzz30903090 rep3070RTX 3070 Ti40904090 repnv 40902272.622658.73208164717841.80841.4023123.7711.6864788750504915973.62421053300113190.0912730.0813154.153.1623232.4213102.751.388.05113405.298.054.312.977.903.175.2823.753.903.1832.493.8610.011.387.947.094.4112.848.1831.8812.9010.204.13.353.1723.513.348.167.072.982269.252640.08208224695836.55839.223396.5911.694794850643918124.0542163327514.0712807.0612808.593.1613145.193.1623390.4413070.811.373.151.377.97112735.27.978.041.3731.654.295.427.034.332.977.853.148.053.335.2323.4910.013.8531.952.9623.423.823.82101.389.874.067.822.9731.7112.773.157.144.4212.987.074.3212.878.278.188.0131.857.8412.7410.015.214.073.3323.523.5683.853.348.217.062.953.163.332269.062638.69208474670836.16839.0123385.4411.6884797150596917444.1143021328124.0812860.5612822.013.1713136.793.1823387.2613063.861.363.141.378.02113115.247.838.031.3831.784.285.237.044.332.997.83.138.143.345.2123.4510.333.883.1731.772.9623.543.893.83101.3910.033.697.932.9731.6612.863.157.074.312.897.064.3112.817.988.27831.797.8812.8110.115.264.093.3223.993.223.547.953.823.3587.12.963.163.338520.028465.82121432346267.74267.437352.8532.8554264543365851814.0835399363288531.9611251.178412.333.1716864.475676.02500.0141.388.10107195.238.024.312.987.853.165.2323.513.853.1732.433.8710.101.387.857.084.3012.858.2332.1212.9510.004.113.353.1723.563.358.177.092.978505.208465.71121682343267.25267.417336.2532.85042651433658519135304370908515.5811231.728397.8016865.295675.99500.0161.38105608.044.317.853.145.223.8410.107.0512.8731.934.083.333.1823.608.102.966827.926800.17105611814214.23214.175959.7526.73856476571101041464.2226593262383.856837.949006.573.156812.523.1513440.974480.59500.011.373.161.388.4575715.697.948.271.4333.474.835.36.974.642.978.153.138.343.45.4824.1911.053.8733.562.9624.124.043.8610.261.3710.254.27.923.1233.3614.343.167.094.3513.077.084.3613.178.58.088.5632.928.0713.3211.056.134.243.424.453.1424.558.653.853.558.347.232.973.153.336824.216795.39105481818213.95213.965956.2426.76956455570941041712.5726638265413.976832.749003.123.156810.553.1613438.44478.41500.0111.373.151.4122.7475745.553.177.948.171.3733.394.715.37.264.8738.963.188.073.923.386.2223.7811.253.343.913.1432.732.957.9623.8224.044.143.8610.3310.341.3810.185.284.064.357.983.0532.3813.1412.893.147.194.3513.088.57.18.384.3213.647.998.38.9832.4233.323.847.149.1517.2310.725.484.073.3524.923.1624.28.043.853.598.367.132.983.162.973.331.386800.66772.9810572210.96213.375978.3856431104298265246810.739036.176838.3213490.244495.98762214780241720.9369738711631322702.6633727346865.693.263.52500.0061.253.31.48.37100615.63.2910.1910.41.2836.555.015.858.166.532.748.753.298.213.833.525.8230.9614.053.434.053.2637.82.9910.4729.0727.434.685.8813.112.961.411.155.884.435.110.33.3938.3313.7715.433.288.334.9915.1110.027.468.465.315.167.999.9410.0836.4238.014.197.2110.1714.6512.095.865.143.4929.124.873.2627.839.054.215.039.888.963.23.293.073.361.4134974557913.1361045561062102110764.4266473658694.23.273.28288.2011.423.311.418.734.98171215.697.713.288.498.441.4434.24.665.654.047.644.194.623.058.423.268.334.283.465.6725.3711.113.463.993.241.4435.073.098.425.0425.484.024.0111.411.4810.811.410.955.614.24.618.423.0834.1313.8613.793.297.664.6913.818.847.588.614.7513.798.458.248.4313.9335.5634.9135.64.067.668.675.928.4513.858.7911.165.74.23.438.4325.6725.13.1258.434.053.418.393.267.733.293.073.283.483.063.441.4335038558313.1361044911062052110584.3468279700684.183.273.28288.1661.453.271.418.414.64172875.687.623.278.588.571.4234.274.655.694.017.594.174.653.068.44.23.298.574.144.683.435.615.648.5226.1110.793.394.093.264.071.4235.283.078.4925.0425.054.024.053.4710.811.0710.841.4210.845.614.094.728.523.0934.2913.5513.683.37.634.6913.558.43.38.727.648.244.6713.678.448.678.4813.7335.0713.7134.110.8633.933.987.558.355.678.5214.038.4211.7625.015.634.213.448.3825.5624.913.283.243.0625.048.4634.223.098.454.043.438.567.673.311.437.863.273.083.443.033.441.4135071558713.1371045281060992107134.1769068678874.193.333.28288.0391.423.31.428.374.71173435.567.273.278.998.371.4234.374.655.893.977.643.754.683.068.423.83.268.754.174.693.475.625.788.3225.0310.913.434.043.274.011.3134.273.138.4326.0825.44.064.043.410.9111.2210.941.4211.55.674.314.678.383.0735.413.6213.953.47.75.2113.698.463.28.387.628.524.6813.658.588.458.4413.5234.1913.6334.2311.2633.94.027.678.255.658.513.68.2610.8225.445.664.23.58.3125.33253.083.262.9825.018.8834.1438.344.223.458.567.273.051.327.623.143.073.313.343.053.511.4235058558413.1261045431059262109914.1667689700404.043.23.25288.0281.393.281.428.384.7171855.638.063.288.378.471.4134.474.675.594.057.354.614.683.048.413.823.298.474.24.663.435.595.778.2925.411.213.433.993.243.951.4134.13.018.4225.2625.164.034.043.3610.9112.511.071.411.095.64.124.688.43.0634.0513.6113.623.237.514.6513.838.463.168.347.638.494.6913.88.18.378.415.2634.3213.4234.111.135.364.017.628.375.748.5513.638.5511.125.265.714.793.469.1926.0925.453.283.063.243.0825.828.3834.472.968.253.953.428.587.253.261.317.553.283.063.273.443.083.371.4220909.0220820.09309454282653.15653.1316886.6610.3991413571439692552074.2155347510054.0421269.7227797.820845.093.1741149.113710.88371.6991.363.141.398.64.322.97144065.217.053.157.828.071.361.3931.894.315.25.233.887.043.834.352.997.874.13.184.33.867.954.034.333.325.275.27.923.4310.053.333.873.183.887.831.3833.012.977.8623.523.553.833.883.3610.3810.0710.110.031.389.975.194.044.317.862.9531.8613.18.0112.883.167.044.312.878.113.168.227.167.994.312.888.018.258.0612.978.231.9412.8233.2210.0331.943.837.057.0423.58.335.217.8314.267.8432.1610.323.585.194.113.398.0323.5123.53.193.162.9423.55832.12.988.0712.863.853.348.387.123.131.393.163.367.523.123.152.993.153.324.082.963.341.3620613.4120517.45311224289648.7116878.210.4281414371439562651714.0854432548144.1120708.8427393.23.1520640.673.1740876.1213606.79371.4221.383.171.372.968.014.32.98144495.297.084.313.197.858.031.371.3831.934.35.25.243.857.094.074.312.977.824.073.174.33.858.094.084.313.335.25.37.9123.4310.073.373.863.197.863.857.821.3731.82.977.8923.4323.523.853.863.3310.049.9810.0610.041.3810.015.224.14.317.92.9632.1112.868.0512.843.175.273.197.084.312.828.043.1612.928.0323.727.068.344.3112.838.258.248.0312.98.1931.9112.8132.0910.0631.9431.973.877.077.0923.478.075.27.8512.777.077.8632.139.9510.2723.45.24.083.368.0523.388.0323.548.063.163.172.9723.488.0131.852.978.0612.893.853.358.027.091.383.181.383.153.367.124.073.323.173.173.842.973.153.344.12.993.361.3922.0648.659.188.069.6724.7453.189.193.988.5517.8111.898.1514.0318.8311.438.3520.7221.113.033.5781.771113.3413.389.8117.756.7110.886.8819.497.127.8110.699.5319.666.9310.598.1312.6812.6419.256.6424.0789.237.5218.89.1918.662.9975.346.0218.655.4255.489.018.997.8123.5923.4423.4823.542.9823.1112.148.6310.0818.256.8773.5129.818.5429.497.2412.137.3416.159.8629.3418.397.2229.3818.2553.4815.8217.889.6228.5917.231817.8228.7318.2470.7628.4171.0822.1970.5370.298.4115.4615.451.2817.6111.31727.6615.3216.9769.4821.522.1550.3211.148.417.0717.0949.717.0649.7517.026.66.436.566.0648.2916.3465.416.0716.5226.337.816.8216.2214.272.695.992.535.975.926.313.27.235.895.465.496.635.095.385.594.484.594.891.7727.1833.944.333.653.7624.8051.713.411.603.129.435.343.256.088.656.253.669.869.351.791.5137.865.556.696.234.378.394.325.253.269.874.183.565.414.559.024.416.174.096.286.229.6828.3612.113.754.723.619.974.749.581.3437.913.379.8428.4028.634.784.534.0212.7312.4212.6012.521.6012.356.574.265.539.693.1138.2915.219.9815.443.665.943.248.295.6715.429.623.9114.648.4227.868.479.195.4915.548.898.839.6215.009.1038.0314.5737.8812.8138.5038.274.608.318.2828.539.056.409.6515.207.459.9038.3212.7313.1527.986.184.263.989.5228.5310.0228.409.143.643.703.623.1029.069.6238.043.2410.0315.564.733.779.077.571.403.442.483.523.833.928.134.143.483.663.694.173.403.763.894.253.343.951.495521480399.2841538961526562903425.4884351814062.933.123.3172.8831.333.461.393.2310.084.673.192037369.815.143.418.8710.551.171.4538.794.947.785.974.159.324.454.643.1210.622.853.34.944.148.134.134.995.185.697.748.3827.7511.395.174.233.538.554.478.911.4238.25311.328.2129.054.094.343.4811.5312.414.1312.981.2714.585.783.944.7210.273.1939.0115.558.4615.444.996.963.367.936.1115.38.963.3616.0510.0927.327.8310.15.1413.979.68.138.9615.9510.0538.7615.8539.0611.7239.3538.624.247.49.2827.319.877.529.9713.687.439.0538.8214.11330.165.814.393.459.1627.4410.5628.5510.113.283.333.344.9328.828.8138.765.199.0415.694.183.558.649.511.353.621.33.364.753.477.864.623.563.485.254.633.183.253.522.823.13.341.435538381198.9621539391559362876515.2781329809994.164.93.3173.0431.413.451.43.1210.755.253.28204046.779.465.343.348.910.231.341.3839.125.275.845.814.419.34.116.793.1810.653.123.315.164.4417.153.965.453.596.018.148.9729.2413.825.274.33.3110.474.3510.181.4137.053.110.3827.5929.124.344.093.4212.1711.5113.0811.241.4513.576.054.165.3310.863.2338.7315.458.8316.393.315.873.357.816.5815.348.373.4415.4110.2329.858.228.645.2315.7210.348.788.7415.48.4537.5913.8838.1712.4738.6539.034.049.349.1629.3510.695.99.2916.799.4410.3938.6913.7310.9630.747.754.593.499.0229.178.2227.258.73.33.33.334.9927.049.5438.795.1110.6116.66.283.478.489.381.423.341.423.443.385.187.984.595.233.363.64.13.153.533.483.913.133.41.465495081328.9671521701551482927684.5184887828754.062.613.6172.8871.073.391.43.18.154.693.1206017.827.026.325.110.018.451.162.9138.994.675.975.844.19.372.645.24.778.73.933.436.545.889.552.816.623.467.446.0710.7529.5413.683.514.373.4710.145.948.851.2638.92.548.6127.7727.894.044.13.3213.6313.1312.4513.461.3313.138.165.925.148.934.6138.5817.310.1516.614.455.584.967.726.1115.629.413.2916.38.2527.259.1110.095.1815.267.7310.178.9115.558.3739.0417.6739.1813.2937.1338.464.127.729.2129.410.037.389.0215.48.268.3538.5811.4113.2527.617.613.933.518.9328.1410.6427.048.343.364.973.263.0729.2910.5438.953.1212.1215.675.263.459.817.481.43.171.424.813.293.379.115.863.433.423.275.824.73.353.54.013.163.171.18OpenBenchmarking.org

vkpeak

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

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20230730int32-scalarcbahgfed3090 rep30904K8K12K16K20KSE +/- 0.34, N = 3SE +/- 0.03, N = 3SE +/- 15.02, N = 32269.062269.252272.626800.606824.216827.928505.208520.0220613.4120909.02

int32-scalar

i: The test quit with a non-zero exit status.

4080: The test quit with a non-zero exit status.

4080 rep: The test quit with a non-zero exit status.

4080 xxx: The test quit with a non-zero exit status.

4080 zzz: The test quit with a non-zero exit status.

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

4090: The test quit with a non-zero exit status.

4090 rep: The test quit with a non-zero exit status.

nv 4090: The test quit with a non-zero exit status.

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20230730int32-vec4cbahgfed3090 rep30904K8K12K16K20KSE +/- 0.26, N = 3SE +/- 0.05, N = 3SE +/- 0.19, N = 32638.692640.082658.736772.986794.926800.178465.718465.8220517.4520820.09

int32-vec4

i: The test quit with a non-zero exit status.

4080: The test quit with a non-zero exit status.

4080 rep: The test quit with a non-zero exit status.

4080 xxx: The test quit with a non-zero exit status.

4080 zzz: The test quit with a non-zero exit status.

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

4090: The test quit with a non-zero exit status.

4090 rep: The test quit with a non-zero exit status.

nv 4090: The test quit with a non-zero exit status.

VkFFT

OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 1.2.31Test: FFT + iFFT C2C 1D batched in double precisiongfhdeiabc30903090 rep40804080 rep4080 zzz4080 xxxnv 409040904090 rep12K24K36K48K60KSE +/- 10.58, N = 3SE +/- 12.42, N = 3SE +/- 11.67, N = 3SE +/- 14.62, N = 31054810561105721214312168147802081620822208473094531122349743503835058350715495055214553831. (CXX) g++ options: -O3

Test: FFT + iFFT C2C 1D batched in double precision

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 1.2.31Test: FFT + iFFT C2C Bluestein benchmark in double precisionfgedi30903090 repcba40804080 rep4080 zzz4080 xxx40904090 repnv 40902K4K6K8K10KSE +/- 11.20, N = 3SE +/- 4.37, N = 3SE +/- 0.33, N = 3181418182343234624174282428946704695471755795583558455878039811981321. (CXX) g++ options: -O3

Test: FFT + iFFT C2C Bluestein benchmark in double precision

3070: The test quit with a non-zero exit status. E: VkFFT System: 2909x2909x1 Buffer: 129 MB avg_time_per_step: 72.294 ms std_error: 0.046 num_iter: 31 benchmark: 1828 bandwidth: 27.9

RTX 3070 Ti: The test quit with a non-zero exit status. E: VkFFT System: 2909x2909x1 Buffer: 129 MB avg_time_per_step: 66.798 ms std_error: 0.291 num_iter: 31 benchmark: 1979 bandwidth: 30.2

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 20230730fp64-vec4hgfed3090cba2004006008001000SE +/- 0.00, N = 3SE +/- 0.48, N = 3SE +/- 0.32, N = 3210.96213.95214.23267.25267.74653.15836.16836.55841.80

fp64-vec4

i: The test quit with a non-zero exit status.

4080: The test quit with a non-zero exit status.

4080 rep: The test quit with a non-zero exit status.

4080 xxx: The test quit with a non-zero exit status.

4080 zzz: The test quit with a non-zero exit status.

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

4090: The test quit with a non-zero exit status.

4090 rep: The test quit with a non-zero exit status.

nv 4090: The test quit with a non-zero exit status.

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20230730fp64-scalarhgfed3090 rep3090cba2004006008001000SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.22, N = 3213.37213.96214.17267.41267.43648.71653.13839.01839.20841.40

fp64-scalar

i: The test quit with a non-zero exit status.

4080: The test quit with a non-zero exit status.

4080 rep: The test quit with a non-zero exit status.

4080 xxx: The test quit with a non-zero exit status.

4080 zzz: The test quit with a non-zero exit status.

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

4090: The test quit with a non-zero exit status.

4090 rep: The test quit with a non-zero exit status.

nv 4090: The test quit with a non-zero exit status.

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20230730int16-vec4gfhed3090 rep3090acb5K10K15K20K25KSE +/- 0.31, N = 3SE +/- 17.33, N = 3SE +/- 21.55, N = 35956.245959.755978.387336.257352.8516878.2016886.6623123.7723385.4423396.59

int16-vec4

i: The test quit with a non-zero exit status.

4080: The test quit with a non-zero exit status.

4080 rep: The test quit with a non-zero exit status.

4080 xxx: The test quit with a non-zero exit status.

4080 zzz: The test quit with a non-zero exit status.

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

4090: The test quit with a non-zero exit status.

4090 rep: The test quit with a non-zero exit status.

nv 4090: The test quit with a non-zero exit status.

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: SingledeRTX 3070 Tigf3070i4080 xxx4080 rep40804080 zzzbca3090 rep30904090nv 40904090 rep816243240SE +/- 0.004, N = 3SE +/- 0.000, N = 3SE +/- 0.029, N = 3SE +/- 0.001, N = 332.85532.85027.18326.76926.73822.06420.93013.13713.13613.13613.12611.69011.68811.68610.42810.3999.2848.9678.9621. (CXX) g++ options: -O3

VkFFT

OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 1.2.31Test: FFT + iFFT C2C 1D batched in single precisiondeabchgfi4080 rep4080 xxx4080 zzz408030903090 repnv 409040904090 rep30K60K90K120K150KSE +/- 2.73, N = 3SE +/- 1.67, N = 3SE +/- 9.54, N = 3SE +/- 25.50, N = 34264542651478874794847971564315645556476697381044911045281045431045561413571414371521701538961539391. (CXX) g++ options: -O3

Test: FFT + iFFT C2C 1D batched in single precision

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 1.2.31Test: FFT + iFFT C2C 1D batched in single precision, no reshufflingdeacbgfi4080 zzz4080 xxx4080 rep40803090 rep30904090nv 40904090 rep30K60K90K120K150KSE +/- 2.33, N = 3SE +/- 2.08, N = 3SE +/- 8.89, N = 343365433655050450596506435709457110711631059261060991062051062101439561439691526561551481559361. (CXX) g++ options: -O3

Test: FFT + iFFT C2C 1D batched in single precision, no reshuffling

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 1.2.31Test: FFT + iFFT C2C 1D batched in half precisiondeacbfghi4080 xxx4080 zzz4080 rep408030903090 rep4090 rep4090nv 409060K120K180K240K300KSE +/- 18.50, N = 3SE +/- 26.03, N = 3SE +/- 83.55, N = 3SE +/- 133.47, N = 385181851919159791744918121041461041711042981322702107132109912110582110762552072651712876512903422927681. (CXX) g++ options: -O3

Test: FFT + iFFT C2C 1D batched in half precision

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: FastestDet307040904090 repnv 409040804080 repf30904080 xxx4080 zzzc3090 repdbRTX 3070 Tiaig246810SE +/- 0.01, N = 3SE +/- 0.23, N = 15SE +/- 0.45, N = 38.655.485.274.514.424.344.224.214.174.164.114.084.084.053.943.622.662.57MIN: 3.94 / MAX: 185.21MIN: 2.67 / MAX: 259.34MIN: 4.05 / MAX: 247.02MIN: 4.34 / MAX: 5.96MIN: 4.25 / MAX: 6.71MIN: 4.19 / MAX: 5.77MIN: 4.18 / MAX: 4.97MIN: 4.19 / MAX: 4.41MIN: 4.05 / MAX: 4.74MIN: 4 / MAX: 4.69MIN: 4.08 / MAX: 4.4MIN: 4.05 / MAX: 4.84MIN: 4.02 / MAX: 4.28MIN: 4.02 / MAX: 4.35MIN: 2.43 / MAX: 267.02MIN: 2.7 / MAX: 4.54MIN: 2.54 / MAX: 3.41MIN: 2.53 / MAX: 3.211. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

VkFFT

OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 1.2.31Test: FFT + iFFT R2C / C2Rhfgiedabc3090 rep309040804080 zzz4080 rep4080 xxx4090 rep4090nv 409020K40K60K80K100KSE +/- 3.71, N = 3SE +/- 118.74, N = 3SE +/- 200.55, N = 3SE +/- 796.66, N = 32652426593266383372735304353994210542163430215443255347664736768968279690688132984351848871. (CXX) g++ options: -O3

Test: FFT + iFFT R2C / C2R

3070: The test quit with a non-zero exit status. E: VkFFT System: 512x512x128 Buffer: 128 MB avg_time_per_step: 4.833 ms std_error: 0.038 num_iter: 31 benchmark: 27226 bandwidth: 311.6

RTX 3070 Ti: The test quit with a non-zero exit status. E: VkFFT System: 512x512x128 Buffer: 128 MB avg_time_per_step: 3.494 ms std_error: 0.002 num_iter: 31 benchmark: 37664 bandwidth: 431.0

OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 1.2.31Test: FFT + iFFT C2C multidimensional in single precisionfgbcaide30903090 rep40804080 xxx4080 zzz4080 rep4090 rep4090nv 409020K40K60K80K100KSE +/- 57.83, N = 3SE +/- 116.12, N = 3SE +/- 437.33, N = 3SE +/- 555.86, N = 326238265413275132812330013468636328370905100554814658696788770040700688099981406828751. (CXX) g++ options: -O3

Test: FFT + iFFT C2C multidimensional in single precision

3070: The test quit with a non-zero exit status. E: VkFFT System: 3840x2160x1 Buffer: 63 MB avg_time_per_step: 2.236 ms std_error: 0.035 num_iter: 64 benchmark: 28982 bandwidth: 331.7

RTX 3070 Ti: The test quit with a non-zero exit status. E: VkFFT System: 3840x2160x1 Buffer: 63 MB avg_time_per_step: 1.462 ms std_error: 0.004 num_iter: 64 benchmark: 44332 bandwidth: 507.3

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: FastestDet3070iRTX 3070 Ti40804080 xxx4080 rep4090 rep3090 repcbnv 409030904080 zzzgf40903691215SE +/- 0.27, N = 149.185.694.334.204.194.184.164.114.084.074.064.044.043.973.852.93MIN: 3.64 / MAX: 122.65MIN: 3.69 / MAX: 261.71MIN: 2.59 / MAX: 433.58MIN: 4.06 / MAX: 4.86MIN: 4.04 / MAX: 5.47MIN: 4.03 / MAX: 5.07MIN: 4 / MAX: 5.58MIN: 4.07 / MAX: 4.21MIN: 4.05 / MAX: 4.36MIN: 4.03 / MAX: 5.83MIN: 3.91 / MAX: 5.78MIN: 4.01 / MAX: 4.15MIN: 3.89 / MAX: 5.01MIN: 3.92 / MAX: 4.75MIN: 3.8 / MAX: 4.65MIN: 2.84 / MAX: 3.381. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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 20230730fp32-scalarhgfedbca3090 rep30905K10K15K20K25KSE +/- 0.30, N = 3SE +/- 16.18, N = 3SE +/- 4.18, N = 36810.736812.996837.948515.588531.9612807.0612860.5613190.0920708.8421269.72

fp32-scalar

i: The test quit with a non-zero exit status.

4080: The test quit with a non-zero exit status.

4080 rep: The test quit with a non-zero exit status.

4080 xxx: The test quit with a non-zero exit status.

4080 zzz: The test quit with a non-zero exit status.

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

4090: The test quit with a non-zero exit status.

4090 rep: The test quit with a non-zero exit status.

nv 4090: The test quit with a non-zero exit status.

OpenBenchmarking.orgGFLOPS, More Is Bettervkpeak 20230730fp32-vec4gfhedabc3090 rep30906K12K18K24K30KSE +/- 2.57, N = 3SE +/- 19.37, N = 3SE +/- 1.81, N = 39002.599006.579036.1711231.7211251.1712730.0812808.5912822.0127393.2027797.80

fp32-vec4

i: The test quit with a non-zero exit status.

4080: The test quit with a non-zero exit status.

4080 rep: The test quit with a non-zero exit status.

4080 xxx: The test quit with a non-zero exit status.

4080 zzz: The test quit with a non-zero exit status.

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

4090: The test quit with a non-zero exit status.

4090 rep: The test quit with a non-zero exit status.

nv 4090: The test quit with a non-zero exit status.

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3 - Model: mobilenet-v330704090 repRTX 3070 Ti4080 xxx4080 rep4080i4080 zzzcb3090 repgf4090nv 4090246810SE +/- 0.18, N = 158.064.903.653.333.273.273.263.203.173.163.153.153.153.122.61MIN: 2.96 / MAX: 219.87MIN: 3.17 / MAX: 120.84MIN: 2.87 / MAX: 347.75MIN: 3.19 / MAX: 4.2MIN: 3.14 / MAX: 3.99MIN: 3.12 / MAX: 5.24MIN: 3.12 / MAX: 4.19MIN: 3.06 / MAX: 3.84MIN: 3.11 / MAX: 8.89MIN: 3.11 / MAX: 3.75MIN: 3.11 / MAX: 3.83MIN: 3.1 / MAX: 3.87MIN: 3.1 / MAX: 3.8MIN: 2.99 / MAX: 5.09MIN: 2.5 / MAX: 3.121. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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 20230730fp16-scalargfhedcba3090 rep30904K8K12K16K20KSE +/- 5.09, N = 3SE +/- 13.46, N = 3SE +/- 4.01, N = 36810.556812.526838.328397.808412.3313136.7913145.1913154.1520640.6720845.09

fp16-scalar

i: The test quit with a non-zero exit status.

4080: The test quit with a non-zero exit status.

4080 rep: The test quit with a non-zero exit status.

4080 xxx: The test quit with a non-zero exit status.

4080 zzz: The test quit with a non-zero exit status.

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

4090: The test quit with a non-zero exit status.

4090 rep: The test quit with a non-zero exit status.

nv 4090: The test quit with a non-zero exit status.

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v2-v2 - Model: mobilenet-v230704090 repRTX 3070 Tinv 4090i40904080 xxx4080 rep40804080 zzzc3090 rep3090dgbaf3691215SE +/- 0.17, N = 15SE +/- 0.02, N = 3SE +/- 0.00, N = 39.674.743.763.603.523.303.283.283.283.253.183.173.173.173.163.163.163.15MIN: 3.19 / MAX: 225.84MIN: 3.09 / MAX: 140.79MIN: 2.6 / MAX: 364.73MIN: 3.43 / MAX: 4.62MIN: 3.29 / MAX: 19.18MIN: 3.11 / MAX: 4.81MIN: 3.1 / MAX: 4.05MIN: 3.11 / MAX: 4MIN: 3.11 / MAX: 3.88MIN: 3.09 / MAX: 4.51MIN: 3.13 / MAX: 3.84MIN: 3.11 / MAX: 4.94MIN: 3.12 / MAX: 4.05MIN: 3.1 / MAX: 8.86MIN: 3.11 / MAX: 3.83MIN: 3.11 / MAX: 3.61MIN: 3.1 / MAX: 3.8MIN: 3.1 / MAX: 3.651. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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 20230730fp16-vec4gfhdeacb3090 rep30909K18K27K36K45KSE +/- 0.37, N = 3SE +/- 0.36, N = 3SE +/- 5.96, N = 313438.4013440.9713490.2416864.4716865.2923232.4223387.2623390.4440876.1241149.10

fp16-vec4

i: The test quit with a non-zero exit status.

4080: The test quit with a non-zero exit status.

4080 rep: The test quit with a non-zero exit status.

4080 xxx: The test quit with a non-zero exit status.

4080 zzz: The test quit with a non-zero exit status.

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

4090: The test quit with a non-zero exit status.

4090 rep: The test quit with a non-zero exit status.

nv 4090: The test quit with a non-zero exit status.

OpenBenchmarking.orgGIOPS, More Is Bettervkpeak 20230730int16-scalargfhedcba3090 rep30903K6K9K12K15KSE +/- 0.02, N = 3SE +/- 0.09, N = 3SE +/- 1.30, N = 34478.414480.594495.985675.995676.0213063.8613070.8113102.7513606.7913710.88

int16-scalar

i: The test quit with a non-zero exit status.

4080: The test quit with a non-zero exit status.

4080 rep: The test quit with a non-zero exit status.

4080 xxx: The test quit with a non-zero exit status.

4080 zzz: The test quit with a non-zero exit status.

3070: The test quit with a non-zero exit status.

RTX 3070 Ti: The test quit with a non-zero exit status.

4090: The test quit with a non-zero exit status.

4090 rep: The test quit with a non-zero exit status.

nv 4090: The test quit with a non-zero exit status.

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: Doubleedgfi30903090 rep40804080 rep4080 xxx4080 zzz4090 repnv 40904090RTX 3070 Ti3070110220330440550SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 3500.02500.01500.01500.01500.01371.70371.42288.20288.17288.04288.03173.04172.89172.8824.8124.751. (CXX) g++ options: -O3

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: blazeface3070RTX 3070 Ti4080 rep4080 xxx40804090 rep4080 zzz3090 repgfb3090c4090inv 40900.71551.4312.14652.8623.5775SE +/- 0.18, N = 153.181.711.451.421.421.411.391.381.381.371.371.361.361.331.251.07MIN: 1.31 / MAX: 185.03MIN: 1.09 / MAX: 448.17MIN: 1.36 / MAX: 8.73MIN: 1.36 / MAX: 1.92MIN: 1.35 / MAX: 2.15MIN: 1.35 / MAX: 1.89MIN: 1.34 / MAX: 1.89MIN: 1.36 / MAX: 1.9MIN: 1.36 / MAX: 1.76MIN: 1.35 / MAX: 1.62MIN: 1.35 / MAX: 1.39MIN: 1.34 / MAX: 1.61MIN: 1.34 / MAX: 1.44MIN: 1.27 / MAX: 1.98MIN: 1.19 / MAX: 2.61MIN: 1.02 / MAX: 1.521. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v2-v2 - Model: mobilenet-v2307040904090 repRTX 3070 Tinv 409040804080 xxxi4080 zzz4080 rep3090 repfgb3090c3691215SE +/- 0.10, N = 159.193.463.453.413.393.313.303.303.283.273.173.163.153.153.143.14MIN: 3.04 / MAX: 232.12MIN: 3.29 / MAX: 4.38MIN: 3.23 / MAX: 4.55MIN: 2.99 / MAX: 184.91MIN: 3.21 / MAX: 4.24MIN: 3.12 / MAX: 4.76MIN: 3.12 / MAX: 4.03MIN: 3.14 / MAX: 4.82MIN: 3.1 / MAX: 4MIN: 3.1 / MAX: 4.34MIN: 3.11 / MAX: 4.5MIN: 3.09 / MAX: 3.89MIN: 3.1 / MAX: 3.63MIN: 3.1 / MAX: 3.68MIN: 3.08 / MAX: 3.7MIN: 3.1 / MAX: 3.671. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: blazeface3070RTX 3070 Ti4080 zzz4080 xxx4080 rep4080gnv 40904090 repi40903090feda3090 repcb0.89551.7912.68653.5824.4775SE +/- 0.16, N = 15SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 33.981.601.421.421.411.411.411.401.401.401.391.391.381.381.381.381.371.371.37MIN: 1.31 / MAX: 228.4MIN: 1.11 / MAX: 436.01MIN: 1.36 / MAX: 2.01MIN: 1.36 / MAX: 1.93MIN: 1.35 / MAX: 1.9MIN: 1.35 / MAX: 2.01MIN: 1.38 / MAX: 2.09MIN: 1.34 / MAX: 1.87MIN: 1.33 / MAX: 1.93MIN: 1.33 / MAX: 2MIN: 1.33 / MAX: 1.94MIN: 1.37 / MAX: 1.82MIN: 1.35 / MAX: 2.08MIN: 1.34 / MAX: 1.88MIN: 1.34 / MAX: 2.25MIN: 1.34 / MAX: 1.85MIN: 1.36 / MAX: 1.46MIN: 1.35 / MAX: 1.82MIN: 1.35 / MAX: 1.751. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: mnasnet307040904090 repRTX 3070 Tinv 40903090 rep246810SE +/- 0.02, N = 38.553.233.123.123.102.96MIN: 2.99 / MAX: 185.5MIN: 3.08 / MAX: 4.73MIN: 3 / MAX: 4.1MIN: 2.97 / MAX: 4.65MIN: 2.97 / MAX: 3.73MIN: 2.92 / MAX: 3.271. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: mobilenetg30704090 rep4090RTX 3070 Ti40803090f4080 rep4080 zzz4080 xxxinv 4090dac3090 repb510152025SE +/- 0.22, N = 15SE +/- 0.06, N = 3SE +/- 0.03, N = 322.7417.8110.7510.089.438.738.608.458.418.388.378.378.158.108.058.028.017.97MIN: 8.24 / MAX: 1264.67MIN: 8.05 / MAX: 159.41MIN: 8.24 / MAX: 287.14MIN: 8.1 / MAX: 118.32MIN: 7.95 / MAX: 398.1MIN: 8.15 / MAX: 10.96MIN: 8.5 / MAX: 13.72MIN: 8.37 / MAX: 9.44MIN: 8.14 / MAX: 11.03MIN: 7.94 / MAX: 10.16MIN: 7.96 / MAX: 9.72MIN: 8.15 / MAX: 9.75MIN: 7.73 / MAX: 9.34MIN: 7.94 / MAX: 14.4MIN: 7.97 / MAX: 9.07MIN: 7.98 / MAX: 8.33MIN: 7.96 / MAX: 9.85MIN: 7.94 / MAX: 8.261. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: alexnet3070RTX 3070 Ti4090 rep40804080 xxx4080 zzznv 409040904080 rep30903090 rep3691215SE +/- 0.16, N = 1511.895.345.254.984.714.704.694.674.644.324.30MIN: 4.34 / MAX: 229.18MIN: 4.25 / MAX: 221.78MIN: 4.86 / MAX: 6.33MIN: 4.59 / MAX: 7.15MIN: 4.26 / MAX: 7.21MIN: 4.28 / MAX: 5.92MIN: 4.28 / MAX: 6.33MIN: 4.28 / MAX: 6MIN: 4.24 / MAX: 6MIN: 4.25 / MAX: 5.33MIN: 4.24 / MAX: 5.111. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: mnasnet30704090 repRTX 3070 Ti4090nv 40903090 rep3090246810SE +/- 0.11, N = 158.153.283.253.193.102.982.97MIN: 2.67 / MAX: 317.68MIN: 3.15 / MAX: 4.32MIN: 2.68 / MAX: 277.21MIN: 3.04 / MAX: 3.98MIN: 2.97 / MAX: 3.92MIN: 2.94 / MAX: 3.36MIN: 2.93 / MAX: 3.451. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

VkFFT

OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 1.2.31Test: FFT + iFFT C2C Bluestein in single precisionfghiedbca30903090 rep40804080 zzz4080 rep4080 xxx40904090 repnv 40904K8K12K16K20KSE +/- 75.16, N = 15SE +/- 72.34, N = 3SE +/- 62.67, N = 3SE +/- 83.38, N = 37571757476221006110560107191127311311113401440614449171211718517287173432037320404206011. (CXX) g++ options: -O3

Test: FFT + iFFT C2C Bluestein in single precision

3070: The test quit with a non-zero exit status. E: VkFFT System: 4241x4241x1 Buffer: 137 MB avg_time_per_step: 16.022 ms std_error: 0.041 num_iter: 29 benchmark: 8770 bandwidth: 133.8

RTX 3070 Ti: The test quit with a non-zero exit status. E: VkFFT System: 4241x4241x1 Buffer: 137 MB avg_time_per_step: 10.602 ms std_error: 0.177 num_iter: 29 benchmark: 13253 bandwidth: 202.2

NCNN

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

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: resnet1830704090 repnv 4090RTX 3070 Ti40904080f4080 rep4080 zzzi4080 xxxg3090 repacd3090b48121620SE +/- 0.17, N = 15SE +/- 0.07, N = 3SE +/- 0.02, N = 314.038.077.826.086.005.695.695.685.635.605.565.555.295.295.245.235.215.20MIN: 5 / MAX: 303.38MIN: 5.86 / MAX: 121.03MIN: 5.54 / MAX: 303.05MIN: 4.97 / MAX: 245.95MIN: 5.47 / MAX: 7.29MIN: 5.16 / MAX: 7.68MIN: 5.22 / MAX: 92.59MIN: 5.17 / MAX: 7.45MIN: 5.08 / MAX: 7.55MIN: 5.13 / MAX: 6.83MIN: 5.09 / MAX: 6.84MIN: 5.19 / MAX: 25.4MIN: 5.18 / MAX: 6.19MIN: 5.09 / MAX: 6.29MIN: 5.15 / MAX: 6.09MIN: 5.1 / MAX: 6.28MIN: 5.09 / MAX: 6.04MIN: 5.1 / MAX: 5.91. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: squeezenet_ssd307040904090 repRTX 3070 Ti4080 zzz40804080 rep4080 xxx3090 rep3090nv 4090510152025SE +/- 0.24, N = 1518.839.819.468.658.067.717.627.277.087.057.02MIN: 6.71 / MAX: 206.11MIN: 7.16 / MAX: 389.1MIN: 7.03 / MAX: 160.39MIN: 6.64 / MAX: 544.17MIN: 7.42 / MAX: 9.25MIN: 7.15 / MAX: 9.1MIN: 7.01 / MAX: 14.37MIN: 6.74 / MAX: 8.84MIN: 7 / MAX: 7.94MIN: 6.97 / MAX: 7.95MIN: 6.38 / MAX: 9.361. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: alexnet3070nv 4090RTX 3070 Ti4090 rep40903090 rep3691215SE +/- 0.57, N = 311.436.326.255.345.144.31MIN: 4.24 / MAX: 178.83MIN: 4.26 / MAX: 195.95MIN: 4.27 / MAX: 334.55MIN: 4.87 / MAX: 6.57MIN: 4.75 / MAX: 7.34MIN: 4.26 / MAX: 4.831. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v2-v2 - Model: mobilenet-v23070nv 4090RTX 3070 Ti40904090 repi4080 zzz40804080 xxx4080 rep3090 repg3090246810SE +/- 0.15, N = 158.355.103.663.413.343.293.283.283.273.273.193.173.15MIN: 3.08 / MAX: 103.38MIN: 3.14 / MAX: 138.88MIN: 3.01 / MAX: 311.25MIN: 3.24 / MAX: 5.42MIN: 3.14 / MAX: 4.45MIN: 3.1 / MAX: 3.96MIN: 3.09 / MAX: 4.98MIN: 3.11 / MAX: 4.16MIN: 3.11 / MAX: 4.73MIN: 3.08 / MAX: 5.18MIN: 3.13 / MAX: 4MIN: 3.1 / MAX: 5.03MIN: 3.1 / MAX: 3.681. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: googlenet3070inv 4090RTX 3070 Ti4080 xxx4090 rep40904080 rep40804080 zzzgbf3090 repc3090510152025SE +/- 0.21, N = 1520.7210.1910.019.868.998.908.878.588.498.378.357.977.947.857.837.82MIN: 7.49 / MAX: 355.33MIN: 7.73 / MAX: 212.36MIN: 7.29 / MAX: 259.11MIN: 7.54 / MAX: 396.21MIN: 8.25 / MAX: 10.27MIN: 8.22 / MAX: 11.07MIN: 8.18 / MAX: 11.09MIN: 7.79 / MAX: 10.48MIN: 7.82 / MAX: 11.98MIN: 7.76 / MAX: 10.31MIN: 8.2 / MAX: 9.39MIN: 7.89 / MAX: 8.7MIN: 7.8 / MAX: 8.78MIN: 7.75 / MAX: 8.64MIN: 7.74 / MAX: 8.61MIN: 7.69 / MAX: 8.61. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: mobilenet30704090i4090 repRTX 3070 Ti4080 rep4080 zzznv 409040804080 xxxfg3090aeb3090 repcd510152025SE +/- 0.25, N = 15SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.00, N = 321.1110.5510.4010.239.358.578.478.458.448.378.278.178.078.058.048.048.038.038.02MIN: 7.98 / MAX: 322.43MIN: 8.22 / MAX: 303.1MIN: 7.97 / MAX: 455.46MIN: 8.13 / MAX: 386.42MIN: 7.49 / MAX: 474.12MIN: 7.98 / MAX: 10MIN: 8.04 / MAX: 10.17MIN: 8.03 / MAX: 12.61MIN: 7.98 / MAX: 10.55MIN: 7.97 / MAX: 16.09MIN: 8.17 / MAX: 9.04MIN: 8.08 / MAX: 9.37MIN: 7.99 / MAX: 8.8MIN: 7.95 / MAX: 8.89MIN: 7.95 / MAX: 9.09MIN: 7.95 / MAX: 14.33MIN: 7.96 / MAX: 8.77MIN: 7.98 / MAX: 8.84MIN: 7.95 / MAX: 9.811. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: blazeface3070RTX 3070 Ti4080f4080 xxx4080 rep4080 zzzc3090 repgb30904090 repi4090nv 40900.68181.36362.04542.72723.409SE +/- 0.19, N = 153.031.791.441.431.421.421.411.381.371.371.371.361.341.281.171.16MIN: 1.28 / MAX: 96.94MIN: 1.13 / MAX: 312.12MIN: 1.37 / MAX: 3.45MIN: 1.4 / MAX: 1.77MIN: 1.36 / MAX: 1.92MIN: 1.36 / MAX: 2.2MIN: 1.34 / MAX: 1.91MIN: 1.36 / MAX: 1.58MIN: 1.35 / MAX: 1.46MIN: 1.34 / MAX: 2.07MIN: 1.35 / MAX: 1.52MIN: 1.34 / MAX: 1.46MIN: 1.27 / MAX: 1.95MIN: 1.23 / MAX: 1.73MIN: 1.11 / MAX: 1.9MIN: 1.11 / MAX: 1.671. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: blazeface3070nv 4090RTX 3070 Ti409030904090 rep3090 rep0.80331.60662.40993.21324.0165SE +/- 0.14, N = 153.572.911.511.451.391.381.38MIN: 1.08 / MAX: 141.04MIN: 1.29 / MAX: 113.97MIN: 1.11 / MAX: 380.46MIN: 1.38 / MAX: 2.98MIN: 1.36 / MAX: 3.12MIN: 1.33 / MAX: 1.98MIN: 1.35 / MAX: 1.881. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: vision_transformer30704090 repnv 40904090RTX 3070 Tii4080 zzz4080 xxx4080 rep4080fg3090 rep3090cb20406080100SE +/- 0.18, N = 1581.7739.1238.9938.7937.8636.5534.4734.3734.2734.2033.4733.3931.9331.8931.7831.65MIN: 44.4 / MAX: 460.28MIN: 33.92 / MAX: 465.83MIN: 34.17 / MAX: 473.06MIN: 33.95 / MAX: 457.41MIN: 32.9 / MAX: 463.9MIN: 33 / MAX: 209.38MIN: 33.32 / MAX: 37.42MIN: 33.01 / MAX: 38.7MIN: 33.07 / MAX: 37.01MIN: 32.92 / MAX: 36.19MIN: 32.89 / MAX: 74.09MIN: 32.73 / MAX: 88.83MIN: 31.76 / MAX: 33.09MIN: 31.66 / MAX: 39.97MIN: 31.64 / MAX: 34.51MIN: 31.53 / MAX: 32.231. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: alexnet3070RTX 3070 Ti4090 repi4090fgnv 40904080 zzz40804080 xxx4080 rep30903090 repbc3691215SE +/- 0.23, N = 1511.005.555.275.014.944.834.714.674.674.664.654.654.314.304.294.28MIN: 4.33 / MAX: 199.92MIN: 4.2 / MAX: 281.58MIN: 4.78 / MAX: 7.7MIN: 4.6 / MAX: 6.68MIN: 4.51 / MAX: 6.64MIN: 4.76 / MAX: 5.74MIN: 4.65 / MAX: 5.57MIN: 4.28 / MAX: 5.7MIN: 4.28 / MAX: 6.29MIN: 4.29 / MAX: 6.1MIN: 4.28 / MAX: 6.42MIN: 4.26 / MAX: 6.13MIN: 4.25 / MAX: 5.13MIN: 4.24 / MAX: 4.99MIN: 4.24 / MAX: 5.64MIN: 4.24 / MAX: 5.121. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: resnet1830704090RTX 3070 Tinv 40904090 rep3090 rep30903691215SE +/- 0.24, N = 1513.347.786.695.975.845.205.20MIN: 5.43 / MAX: 279.86MIN: 5.4 / MAX: 168.29MIN: 5.06 / MAX: 462.37MIN: 5.4 / MAX: 8.25MIN: 5.35 / MAX: 8.28MIN: 5.08 / MAX: 6.05MIN: 5.1 / MAX: 6.051. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: resnet183070RTX 3070 Ti40904080 xxxinv 40904090 rep4080 rep40804080 zzzgbf3090 rep3090c3691215SE +/- 0.16, N = 1513.386.235.975.895.855.845.815.695.655.595.505.425.305.245.235.23MIN: 5.43 / MAX: 208.42MIN: 4.99 / MAX: 309.18MIN: 5.46 / MAX: 7.02MIN: 5.36 / MAX: 7.53MIN: 5.3 / MAX: 8.27MIN: 5.35 / MAX: 7.72MIN: 5.3 / MAX: 6.82MIN: 5.11 / MAX: 6.94MIN: 5.14 / MAX: 6.93MIN: 5.09 / MAX: 7.7MIN: 5.4 / MAX: 6.38MIN: 5.36 / MAX: 6.27MIN: 5.17 / MAX: 5.93MIN: 5.14 / MAX: 5.99MIN: 5.1 / MAX: 6.07MIN: 5.11 / MAX: 6.031. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: efficientnet-b030704090 repRTX 3070 Ti4090nv 40904080 zzz40804080 rep4080 xxx30903090 rep3691215SE +/- 0.13, N = 159.814.414.374.154.104.054.044.013.973.883.85MIN: 3.87 / MAX: 165.38MIN: 4.21 / MAX: 5.82MIN: 3.85 / MAX: 366.28MIN: 3.93 / MAX: 5.94MIN: 3.87 / MAX: 6.14MIN: 3.83 / MAX: 5.42MIN: 3.84 / MAX: 4.83MIN: 3.81 / MAX: 6.04MIN: 3.79 / MAX: 5.93MIN: 3.83 / MAX: 4.72MIN: 3.78 / MAX: 4.831. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: squeezenet_ssd3070nv 409040904090 repRTX 3070 Tii4080 xxx40804080 rep4080 zzzg3090 rep3090cbf48121620SE +/- 0.24, N = 1517.759.379.329.308.398.167.647.647.597.357.267.097.047.047.036.97MIN: 6.47 / MAX: 272.11MIN: 7.07 / MAX: 281.92MIN: 7.1 / MAX: 172.56MIN: 6.92 / MAX: 310.91MIN: 6.53 / MAX: 436.05MIN: 7.51 / MAX: 9.94MIN: 7.03 / MAX: 9.19MIN: 7.05 / MAX: 9.9MIN: 7.02 / MAX: 8.87MIN: 6.79 / MAX: 9.82MIN: 7.14 / MAX: 8.59MIN: 7.02 / MAX: 7.99MIN: 6.96 / MAX: 7.74MIN: 6.96 / MAX: 7.83MIN: 6.97 / MAX: 7.88MIN: 6.83 / MAX: 13.871. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: FastestDet30704080 zzz4090RTX 3070 Ti40804080 rep4090 rep3090 rep30904080 xxxnv 4090246810SE +/- 0.27, N = 156.714.614.454.324.194.174.114.073.833.752.64MIN: 2.73 / MAX: 109.52MIN: 4.45 / MAX: 5.92MIN: 4.29 / MAX: 5.05MIN: 2.51 / MAX: 398.91MIN: 4.06 / MAX: 7.41MIN: 4.02 / MAX: 4.75MIN: 3.98 / MAX: 4.73MIN: 4.03 / MAX: 4.18MIN: 3.79 / MAX: 4.09MIN: 3.63 / MAX: 5.24MIN: 2.52 / MAX: 4.141. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: alexnet30704090 repiRTX 3070 Tinv 4090g4080 zzz4080 xxx4080 rep4090f40803090cb3090 repeda3691215SE +/- 0.18, N = 15SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 310.886.796.535.255.204.874.684.684.654.644.644.624.354.334.334.314.314.314.31MIN: 4.38 / MAX: 52.99MIN: 4.23 / MAX: 262.43MIN: 4.57 / MAX: 242.16MIN: 4.23 / MAX: 375.94MIN: 4.82 / MAX: 7.07MIN: 4.8 / MAX: 5.62MIN: 4.26 / MAX: 6.23MIN: 4.26 / MAX: 6.61MIN: 4.26 / MAX: 6.53MIN: 4.26 / MAX: 5.98MIN: 4.57 / MAX: 5.49MIN: 4.26 / MAX: 6.15MIN: 4.28 / MAX: 7.49MIN: 4.26 / MAX: 10.59MIN: 4.28 / MAX: 5.16MIN: 4.26 / MAX: 5.26MIN: 4.23 / MAX: 11.03MIN: 4.25 / MAX: 5.28MIN: 4.24 / MAX: 5.21. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: mnasnet3070nv 4090RTX 3070 Ti4090 rep40904080 xxx4080 rep40804080 zzzg3090cd3090 repfbai246810SE +/- 0.14, N = 15SE +/- 0.01, N = 3SE +/- 0.01, N = 36.884.773.263.223.123.063.063.053.043.002.992.992.982.972.972.972.972.74MIN: 3.05 / MAX: 110.25MIN: 3.07 / MAX: 97.57MIN: 2.46 / MAX: 277.54MIN: 3.11 / MAX: 3.71MIN: 2.98 / MAX: 3.79MIN: 2.94 / MAX: 4.45MIN: 2.94 / MAX: 4.51MIN: 2.92 / MAX: 3.82MIN: 2.91 / MAX: 4.47MIN: 2.96 / MAX: 3.68MIN: 2.95 / MAX: 3.88MIN: 2.96 / MAX: 3.44MIN: 2.94 / MAX: 3.83MIN: 2.93 / MAX: 3.28MIN: 2.93 / MAX: 3.95MIN: 2.93 / MAX: 3.45MIN: 2.92 / MAX: 3.48MIN: 2.62 / MAX: 4.221. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: googlenet30704090 rep4090RTX 3070 Tiginv 40904080 xxx40804080 zzz4080 repfa3090edb3090 repc510152025SE +/- 0.22, N = 15SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.04, N = 319.4910.6510.629.878.968.758.708.428.428.418.408.157.907.877.857.857.857.827.80MIN: 7.4 / MAX: 200.01MIN: 8.29 / MAX: 236.11MIN: 7.83 / MAX: 323.31MIN: 7.33 / MAX: 399.24MIN: 8.82 / MAX: 9.87MIN: 8.08 / MAX: 16.01MIN: 7.96 / MAX: 10.01MIN: 7.73 / MAX: 10.06MIN: 7.79 / MAX: 10.01MIN: 7.72 / MAX: 9.9MIN: 7.77 / MAX: 9.78MIN: 8.02 / MAX: 9.02MIN: 7.74 / MAX: 9.54MIN: 7.76 / MAX: 10.36MIN: 7.71 / MAX: 8.76MIN: 7.71 / MAX: 8.85MIN: 7.76 / MAX: 8.76MIN: 7.69 / MAX: 8.61MIN: 7.72 / MAX: 8.741. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: FastestDet30704080 repRTX 3070 Ti30903090 repnv 40904080 zzz4080 xxx4090 rep4090246810SE +/- 0.87, N = 37.124.204.184.104.073.933.823.803.122.85MIN: 3.72 / MAX: 188.7MIN: 4.04 / MAX: 5.63MIN: 2.53 / MAX: 295.11MIN: 4.07 / MAX: 4.34MIN: 4.03 / MAX: 4.2MIN: 3.76 / MAX: 11.77MIN: 3.65 / MAX: 9.77MIN: 3.65 / MAX: 6.08MIN: 2.97 / MAX: 4.42MIN: 2.74 / MAX: 4.361. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v2-v2 - Model: mobilenet-v23070RTX 3070 Tinv 40904090 rep40904080 zzz4080 repi4080 xxx40803090g3090 repadebfc246810SE +/- 0.14, N = 15SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 37.813.563.433.313.303.293.293.293.263.263.183.183.173.173.163.143.143.133.13MIN: 3.07 / MAX: 154.75MIN: 3.09 / MAX: 345.01MIN: 3.25 / MAX: 4.81MIN: 3.14 / MAX: 4.92MIN: 3.12 / MAX: 4.82MIN: 3.11 / MAX: 3.98MIN: 3.12 / MAX: 4.14MIN: 3.12 / MAX: 3.93MIN: 3.1 / MAX: 3.87MIN: 3.1 / MAX: 4.12MIN: 3.14 / MAX: 3.63MIN: 3.13 / MAX: 3.9MIN: 3.12 / MAX: 3.64MIN: 3.09 / MAX: 3.78MIN: 3.09 / MAX: 3.92MIN: 3.08 / MAX: 4.06MIN: 3.1 / MAX: 3.73MIN: 3.07 / MAX: 3.82MIN: 3.08 / MAX: 3.851. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: alexnet3070nv 4090RTX 3070 Ti4090 rep40903090 rep30903691215SE +/- 0.21, N = 1510.696.545.415.164.944.304.30MIN: 4.32 / MAX: 148.92MIN: 4.56 / MAX: 110.58MIN: 4.23 / MAX: 364.66MIN: 4.73 / MAX: 6.38MIN: 4.52 / MAX: 6.23MIN: 4.24 / MAX: 4.85MIN: 4.25 / MAX: 4.631. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: efficientnet-b03070nv 4090RTX 3070 Ti4090 rep409030903090 rep3691215SE +/- 0.16, N = 159.535.884.554.444.143.863.85MIN: 3.77 / MAX: 182.53MIN: 3.96 / MAX: 194.08MIN: 3.84 / MAX: 379.07MIN: 4.24 / MAX: 5.18MIN: 3.93 / MAX: 5.94MIN: 3.82 / MAX: 4.82MIN: 3.81 / MAX: 4.61. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: regnety_400m30704090 repnv 4090RTX 3070 Ti4080 xxx4080 rep4080 zzzf4080ic40903090 repgb3090510152025SE +/- 0.21, N = 1519.6617.159.559.028.758.578.478.348.338.218.148.138.098.078.057.95MIN: 7.5 / MAX: 235.36MIN: 8.02 / MAX: 773.45MIN: 7.5 / MAX: 193.79MIN: 7.69 / MAX: 501.76MIN: 8.35 / MAX: 10.08MIN: 8.21 / MAX: 10.39MIN: 8.13 / MAX: 10.27MIN: 8.26 / MAX: 9.3MIN: 8.02 / MAX: 9.64MIN: 7.9 / MAX: 9.99MIN: 8.08 / MAX: 8.69MIN: 7.75 / MAX: 10.05MIN: 7.99 / MAX: 14.25MIN: 7.97 / MAX: 8.81MIN: 8 / MAX: 8.58MIN: 7.88 / MAX: 8.671. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: FastestDet3070RTX 3070 Ti40804080 zzz4080 xxx4080 rep40903090 rep30904090 repginv 4090246810SE +/- 0.20, N = 156.934.414.284.204.174.144.134.084.033.963.923.832.81MIN: 2.57 / MAX: 163.84MIN: 2.06 / MAX: 295.24MIN: 4.13 / MAX: 4.85MIN: 4.01 / MAX: 11.47MIN: 4.03 / MAX: 5.63MIN: 4 / MAX: 5.6MIN: 3.99 / MAX: 4.67MIN: 4.04 / MAX: 4.29MIN: 3.99 / MAX: 4.22MIN: 3.79 / MAX: 11.36MIN: 3.88 / MAX: 4.72MIN: 3.7 / MAX: 4.57MIN: 2.68 / MAX: 4.381. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: alexnet3070nv 4090RTX 3070 Ti4090 rep40904080 xxx4080 rep4080 zzz30903090 rep3691215SE +/- 0.43, N = 310.596.626.175.454.994.694.684.664.334.31MIN: 4.3 / MAX: 177.68MIN: 4.28 / MAX: 339.62MIN: 4.5 / MAX: 261.75MIN: 4.93 / MAX: 7.98MIN: 4.56 / MAX: 6.91MIN: 4.26 / MAX: 6.07MIN: 4.27 / MAX: 6.08MIN: 4.24 / MAX: 5.97MIN: 4.26 / MAX: 5.19MIN: 4.26 / MAX: 5.071. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: shufflenet-v230704090RTX 3070 Ti4090 repi4080 xxxnv 409040804080 zzz4080 repfgc3090 repb3090246810SE +/- 0.21, N = 158.135.184.093.593.523.473.463.463.433.433.403.383.343.333.333.32MIN: 3.09 / MAX: 147.21MIN: 3.34 / MAX: 283.54MIN: 3.12 / MAX: 435.28MIN: 3.46 / MAX: 4.09MIN: 3.39 / MAX: 4.05MIN: 3.33 / MAX: 5.01MIN: 3.32 / MAX: 5.2MIN: 3.34 / MAX: 3.93MIN: 3.31 / MAX: 3.94MIN: 3.3 / MAX: 4.03MIN: 3.35 / MAX: 4.17MIN: 3.34 / MAX: 4.15MIN: 3.32 / MAX: 3.79MIN: 3.3 / MAX: 3.67MIN: 3.3 / MAX: 3.79MIN: 3.28 / MAX: 3.661. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: resnet183070nv 4090RTX 3070 Tig4090 repi409040804080 xxx4080 rep4080 zzzfa3090dbec3090 rep3691215SE +/- 0.20, N = 15SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 312.687.446.286.226.015.825.695.675.625.615.595.485.285.275.235.235.225.215.20MIN: 5.39 / MAX: 262.62MIN: 5.29 / MAX: 320.54MIN: 4.94 / MAX: 298.06MIN: 6.11 / MAX: 7MIN: 5.44 / MAX: 8.18MIN: 5.28 / MAX: 7.02MIN: 5.16 / MAX: 8.22MIN: 5.18 / MAX: 7.22MIN: 5.1 / MAX: 7.65MIN: 5.11 / MAX: 7.44MIN: 5.06 / MAX: 6.95MIN: 5.33 / MAX: 6.16MIN: 5.17 / MAX: 6.16MIN: 5.15 / MAX: 6.19MIN: 5.08 / MAX: 6.28MIN: 5.13 / MAX: 6.18MIN: 5.09 / MAX: 11.15MIN: 5.11 / MAX: 6.04MIN: 5.09 / MAX: 5.981. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: resnet1830704090 rep4090RTX 3070 Tinv 40904080 xxx4080 zzz4080 rep3090 rep30903691215SE +/- 0.30, N = 312.648.147.746.226.075.785.775.645.305.20MIN: 5.3 / MAX: 53.81MIN: 5.39 / MAX: 122.47MIN: 5.25 / MAX: 312.09MIN: 5.3 / MAX: 8.22MIN: 5.49 / MAX: 15.12MIN: 5.21 / MAX: 6.97MIN: 5.22 / MAX: 7.06MIN: 5.11 / MAX: 7.51MIN: 5.21 / MAX: 6.24MIN: 5.1 / MAX: 6.161. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: googlenet3070nv 4090RTX 3070 Ti4090 rep4080 rep40904080 xxx4080 zzz3090 rep3090510152025SE +/- 0.80, N = 319.2010.759.688.978.528.388.328.297.917.90MIN: 7.84 / MAX: 193.36MIN: 7.92 / MAX: 447.83MIN: 8.16 / MAX: 382.41MIN: 8.22 / MAX: 10.51MIN: 7.85 / MAX: 10.56MIN: 7.78 / MAX: 10.43MIN: 7.71 / MAX: 10.39MIN: 7.63 / MAX: 9.87MIN: 7.81 / MAX: 8.62MIN: 7.8 / MAX: 8.731. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: vgg163070inv 40904090 repRTX 3070 Ti40904080 rep4080 zzz40804080 xxxfgadbc3090 rep30901326395265SE +/- 0.23, N = 15SE +/- 0.30, N = 3SE +/- 0.05, N = 356.6430.9629.5429.2428.3627.7526.1125.4025.3725.0324.1923.7823.7523.5123.4923.4523.4323.43MIN: 25.75 / MAX: 367.74MIN: 25.92 / MAX: 328.63MIN: 24.77 / MAX: 364.86MIN: 26.51 / MAX: 270.71MIN: 24.13 / MAX: 449.57MIN: 24.58 / MAX: 282.59MIN: 24.54 / MAX: 30.29MIN: 24.09 / MAX: 32.86MIN: 24.26 / MAX: 36.52MIN: 23.85 / MAX: 28.9MIN: 23.99 / MAX: 30.98MIN: 23.52 / MAX: 24.89MIN: 23.31 / MAX: 25.12MIN: 23.19 / MAX: 24.68MIN: 23.36 / MAX: 24.62MIN: 23.26 / MAX: 24.51MIN: 23.23 / MAX: 24.39MIN: 23.2 / MAX: 24.11. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: resnet503070i4090 repnv 4090RTX 3070 Ti4090g4080 zzz4080f4080 xxx4080 repc3090 rep3090b612182430SE +/- 0.22, N = 1524.0714.0513.8213.6812.1111.3911.2511.2111.1111.0510.9110.7910.3310.0710.0510.01MIN: 10.02 / MAX: 218.35MIN: 11.69 / MAX: 252.21MIN: 10.34 / MAX: 245.6MIN: 10.25 / MAX: 566.67MIN: 10.16 / MAX: 382.56MIN: 10.48 / MAX: 13.29MIN: 10.55 / MAX: 118.12MIN: 10.3 / MAX: 13.25MIN: 10.19 / MAX: 13.03MIN: 10.46 / MAX: 112.6MIN: 9.91 / MAX: 13.1MIN: 9.91 / MAX: 12.75MIN: 10.16 / MAX: 13.97MIN: 9.94 / MAX: 11.06MIN: 9.85 / MAX: 12.64MIN: 9.89 / MAX: 10.861. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: shufflenet-v230704090 rep4090RTX 3070 Tinv 409040804080 zzz4080 xxxi4080 rep3090 repg3090246810SE +/- 0.16, N = 158.005.275.173.753.513.463.433.433.433.393.373.343.33MIN: 3.16 / MAX: 190.15MIN: 3.27 / MAX: 191.55MIN: 3.22 / MAX: 208.13MIN: 3.2 / MAX: 361.52MIN: 3.38 / MAX: 4.05MIN: 3.3 / MAX: 5.74MIN: 3.29 / MAX: 3.87MIN: 3.31 / MAX: 3.95MIN: 3.3 / MAX: 4.89MIN: 3.26 / MAX: 3.91MIN: 3.33 / MAX: 3.8MIN: 3.31 / MAX: 4.05MIN: 3.29 / MAX: 3.671. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: efficientnet-b03070RTX 3070 Tinv 40904090 rep40904080 repi4080 xxx4080 zzz4080gac3090f3090 repdb3691215SE +/- 0.19, N = 15SE +/- 0.05, N = 3SE +/- 0.00, N = 39.234.724.374.304.234.094.054.043.993.993.913.903.883.873.873.863.853.85MIN: 3.43 / MAX: 156.19MIN: 3.37 / MAX: 486.93MIN: 4.15 / MAX: 5.96MIN: 4.08 / MAX: 5.07MIN: 3.98 / MAX: 12.23MIN: 3.86 / MAX: 5.59MIN: 3.78 / MAX: 5.45MIN: 3.83 / MAX: 5.71MIN: 3.8 / MAX: 5.69MIN: 3.79 / MAX: 5.83MIN: 3.85 / MAX: 4.64MIN: 3.82 / MAX: 4.51MIN: 3.84 / MAX: 4.41MIN: 3.83 / MAX: 4.69MIN: 3.81 / MAX: 4.97MIN: 3.81 / MAX: 4.75MIN: 3.81 / MAX: 4.46MIN: 3.81 / MAX: 4.421. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3 - Model: mobilenet-v33070RTX 3070 Ti4090nv 40904090 rep4080 xxx4080 repi4080 zzz40803090 rep3090adcg246810SE +/- 0.21, N = 14SE +/- 0.00, N = 2SE +/- 0.00, N = 37.523.613.533.473.413.273.263.263.243.243.193.183.183.173.173.14MIN: 2.94 / MAX: 215MIN: 2.51 / MAX: 502.85MIN: 3.39 / MAX: 4.31MIN: 3.32 / MAX: 4.91MIN: 3.27 / MAX: 5.24MIN: 3.13 / MAX: 3.85MIN: 3.09 / MAX: 3.96MIN: 3.14 / MAX: 3.9MIN: 3.11 / MAX: 4.47MIN: 3.09 / MAX: 4.73MIN: 3.15 / MAX: 3.72MIN: 3.14 / MAX: 4.14MIN: 3.14 / MAX: 3.82MIN: 3.12 / MAX: 3.96MIN: 3.15 / MAX: 3.74MIN: 3.1 / MAX: 3.811. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: googlenet30704090 repnv 4090RTX 3070 Ti40903090 rep510152025SE +/- 0.55, N = 318.8010.4710.149.978.557.86MIN: 7.78 / MAX: 141.46MIN: 7.86 / MAX: 191.94MIN: 7.85 / MAX: 257.61MIN: 8.16 / MAX: 381.49MIN: 7.85 / MAX: 11.39MIN: 7.75 / MAX: 8.571. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: efficientnet-b03070nv 4090RTX 3070 Ti40904090 rep4080 rep4080 xxx4080 zzz30903090 rep3691215SE +/- 0.46, N = 39.195.944.744.474.354.074.013.953.883.85MIN: 3.85 / MAX: 131.42MIN: 3.97 / MAX: 208.59MIN: 3.68 / MAX: 295.7MIN: 4.23 / MAX: 5.82MIN: 4.08 / MAX: 5.62MIN: 3.85 / MAX: 4.79MIN: 3.83 / MAX: 5.28MIN: 3.79 / MAX: 4.59MIN: 3.83 / MAX: 4.61MIN: 3.81 / MAX: 4.751. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: googlenet30704090 repRTX 3070 Ti4090nv 409030903090 rep510152025SE +/- 0.22, N = 1518.6610.189.588.918.857.837.82MIN: 7.42 / MAX: 326.73MIN: 7.81 / MAX: 204.67MIN: 7.62 / MAX: 396.9MIN: 8.3 / MAX: 10.96MIN: 8.16 / MAX: 10.25MIN: 7.73 / MAX: 8.6MIN: 7.72 / MAX: 8.61. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: blazeface3070408040904080 rep4090 rep4080 zzz30903090 repRTX 3070 Ti4080 xxxnv 40900.67281.34562.01842.69123.364SE +/- 0.03, N = 152.991.441.421.421.411.411.381.371.341.311.26MIN: 1.22 / MAX: 149.55MIN: 1.37 / MAX: 2.07MIN: 1.36 / MAX: 1.92MIN: 1.35 / MAX: 2.89MIN: 1.35 / MAX: 1.91MIN: 1.34 / MAX: 1.88MIN: 1.36 / MAX: 1.53MIN: 1.35 / MAX: 1.48MIN: 1.06 / MAX: 2.66MIN: 1.25 / MAX: 3.14MIN: 1.2 / MAX: 1.761. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: vision_transformer3070nv 40904090RTX 3070 Ti4090 repi4080 rep40804080 xxx4080 zzzf3090gadb3090 repc20406080100SE +/- 0.12, N = 15SE +/- 0.29, N = 3SE +/- 0.39, N = 375.3438.9038.2537.9137.8137.8035.2835.0734.2734.1033.5633.0132.7332.4932.4331.9531.8031.77MIN: 38.72 / MAX: 418.01MIN: 34.2 / MAX: 300.84MIN: 33.04 / MAX: 447.7MIN: 32.08 / MAX: 541.11MIN: 32.66 / MAX: 453.44MIN: 33.74 / MAX: 321.51MIN: 33.9 / MAX: 38.67MIN: 33.14 / MAX: 43.26MIN: 32.82 / MAX: 39.79MIN: 32.65 / MAX: 37.64MIN: 32.98 / MAX: 51.93MIN: 32.88 / MAX: 33.42MIN: 31.44 / MAX: 81.32MIN: 31.67 / MAX: 40.11MIN: 31.56 / MAX: 37.69MIN: 31.79 / MAX: 32.33MIN: 31.66 / MAX: 32.23MIN: 31.61 / MAX: 35.681. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: mnasnet3070RTX 3070 Ti4080 xxx4090 rep40804080 rep4080 zzz4090ig3090 rep3090fcbnv 4090246810SE +/- 0.16, N = 146.023.373.133.103.093.073.013.002.992.982.972.972.962.962.962.54MIN: 2.79 / MAX: 50.49MIN: 2.86 / MAX: 278.87MIN: 3 / MAX: 5.1MIN: 2.97 / MAX: 3.72MIN: 2.94 / MAX: 3.79MIN: 2.94 / MAX: 3.72MIN: 2.91 / MAX: 3.6MIN: 2.89 / MAX: 3.46MIN: 2.86 / MAX: 4.38MIN: 2.95 / MAX: 3.63MIN: 2.94 / MAX: 3.39MIN: 2.92 / MAX: 3.28MIN: 2.92 / MAX: 3.81MIN: 2.93 / MAX: 3.41MIN: 2.93 / MAX: 3.4MIN: 2.44 / MAX: 3.581. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: googlenet30704090i4090 repRTX 3070 Tinv 40904080 rep4080 xxx4080 zzz4080g3090 rep3090510152025SE +/- 0.24, N = 1518.6011.3010.4710.389.848.618.498.438.428.407.967.897.86MIN: 8.02 / MAX: 292.16MIN: 7.95 / MAX: 477.54MIN: 8.21 / MAX: 350.07MIN: 7.96 / MAX: 255.68MIN: 7.3 / MAX: 438.04MIN: 7.95 / MAX: 10.07MIN: 7.74 / MAX: 10.76MIN: 7.77 / MAX: 10.4MIN: 7.78 / MAX: 10.7MIN: 7.71 / MAX: 10.64MIN: 7.81 / MAX: 9.05MIN: 7.79 / MAX: 8.84MIN: 7.74 / MAX: 8.621. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: vgg163070iRTX 3070 Ti4090nv 40904090 rep4080 xxx4080 zzz4080 rep4080gfc30903090 repb1224364860SE +/- 0.27, N = 1555.4229.0728.4028.2127.7727.5926.0825.2625.0425.0424.7124.1223.5423.5023.4323.42MIN: 25.32 / MAX: 281.46MIN: 24.45 / MAX: 263.33MIN: 23.98 / MAX: 456MIN: 24.57 / MAX: 270.76MIN: 24.82 / MAX: 264.66MIN: 24.34 / MAX: 396.09MIN: 24.52 / MAX: 27.73MIN: 24.14 / MAX: 27.73MIN: 23.81 / MAX: 27.15MIN: 23.87 / MAX: 28.04MIN: 23.88 / MAX: 119.23MIN: 23.57 / MAX: 46.44MIN: 23.32 / MAX: 24.54MIN: 23.23 / MAX: 24.26MIN: 23.26 / MAX: 24.3MIN: 23.27 / MAX: 24.321. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: vgg16307040904090 repRTX 3070 Tinv 4090i40804080 xxx4080 zzz4080 repg30903090 rep1224364860SE +/- 0.30, N = 1555.4831.5729.1228.6327.8927.4325.4825.4025.1625.0524.0423.5523.52MIN: 25.94 / MAX: 298.67MIN: 26.09 / MAX: 318.58MIN: 24.62 / MAX: 266.39MIN: 24.13 / MAX: 500.18MIN: 24.5 / MAX: 463.23MIN: 24.65 / MAX: 251.37MIN: 23.88 / MAX: 51.68MIN: 24.05 / MAX: 27.09MIN: 23.97 / MAX: 27.81MIN: 23.78 / MAX: 26.95MIN: 23.48 / MAX: 73.3MIN: 23.31 / MAX: 24.48MIN: 23.33 / MAX: 25.081. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: efficientnet-b03070RTX 3070 Tii4090 repg40904080 xxxnv 4090f4080 zzz4080 rep4080c3090 rep3090b3691215SE +/- 0.22, N = 159.014.784.684.344.144.094.064.044.044.034.024.023.893.853.833.82MIN: 3.98 / MAX: 188.57MIN: 3.82 / MAX: 411.19MIN: 4.48 / MAX: 6.02MIN: 4.16 / MAX: 5.28MIN: 4.09 / MAX: 5.13MIN: 3.86 / MAX: 4.83MIN: 3.83 / MAX: 5.55MIN: 3.78 / MAX: 4.9MIN: 3.99 / MAX: 4.82MIN: 3.82 / MAX: 5.43MIN: 3.82 / MAX: 5.39MIN: 3.82 / MAX: 5.66MIN: 3.83 / MAX: 9.72MIN: 3.81 / MAX: 4.53MIN: 3.78 / MAX: 4.41MIN: 3.79 / MAX: 4.341. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: efficientnet-b03070iRTX 3070 Ti4090nv 40904090 rep4080 rep4080 zzz4080 xxx40803090d3090 repgfaecb3691215SE +/- 0.18, N = 15SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 38.995.884.534.344.104.094.054.044.044.013.883.873.863.863.863.863.843.833.82MIN: 3.71 / MAX: 129.99MIN: 4.04 / MAX: 364.21MIN: 3.75 / MAX: 396.62MIN: 4.14 / MAX: 5.84MIN: 3.86 / MAX: 5.46MIN: 3.87 / MAX: 5.46MIN: 3.83 / MAX: 6.11MIN: 3.8 / MAX: 5.31MIN: 3.82 / MAX: 5.33MIN: 3.78 / MAX: 5.34MIN: 3.84 / MAX: 4.39MIN: 3.77 / MAX: 9.91MIN: 3.82 / MAX: 4.34MIN: 3.82 / MAX: 4.22MIN: 3.78 / MAX: 10.45MIN: 3.8 / MAX: 4.6MIN: 3.79 / MAX: 4.76MIN: 3.79 / MAX: 4.61MIN: 3.78 / MAX: 4.391. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: shufflenet-v23070RTX 3070 Ti40904080 rep4090 rep4080 xxx30904080 zzz3090 repnv 4090246810SE +/- 0.60, N = 37.814.023.483.473.423.403.363.363.333.32MIN: 3.3 / MAX: 131.26MIN: 3.27 / MAX: 328.59MIN: 3.35 / MAX: 4.05MIN: 3.33 / MAX: 5.39MIN: 3.29 / MAX: 3.94MIN: 3.28 / MAX: 3.87MIN: 3.32 / MAX: 3.66MIN: 3.23 / MAX: 3.99MIN: 3.3 / MAX: 3.78MIN: 3.19 / MAX: 4.761. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: resnet5030704090nv 4090iRTX 3070 Ti4090 rep40804080 zzz4080 xxx4080 rep3090g3090 rep612182430SE +/- 0.23, N = 1523.5914.0813.6313.1012.7312.1711.4010.9110.9110.8010.3810.3310.04MIN: 9.96 / MAX: 177.63MIN: 10.29 / MAX: 247.29MIN: 10.52 / MAX: 488.94MIN: 10.59 / MAX: 267.95MIN: 9.84 / MAX: 518.97MIN: 11.25 / MAX: 13.79MIN: 10.5 / MAX: 13.51MIN: 9.94 / MAX: 14.83MIN: 9.91 / MAX: 13.07MIN: 9.89 / MAX: 12.54MIN: 9.88 / MAX: 18.75MIN: 10.2 / MAX: 11.18MIN: 9.94 / MAX: 10.891. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: resnet503070nv 40904080 zzzRTX 3070 Ti40904090 rep40804080 xxx4080 rep30903090 rep612182430SE +/- 0.25, N = 1523.4413.1312.5012.4212.4011.5111.4811.2211.0710.079.98MIN: 10.17 / MAX: 219.36MIN: 10.18 / MAX: 247.5MIN: 11.47 / MAX: 14.56MIN: 10.23 / MAX: 444.76MIN: 11.44 / MAX: 14.43MIN: 10.56 / MAX: 13.22MIN: 10.56 / MAX: 12.93MIN: 10.33 / MAX: 12.81MIN: 10.16 / MAX: 13.16MIN: 9.95 / MAX: 10.88MIN: 9.85 / MAX: 11.351. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: resnet50307040904090 repiRTX 3070 Tinv 40904080 zzz4080 xxx4080 rep4080gf3090ed3090 repacb612182430SE +/- 0.26, N = 15SE +/- 0.12, N = 3SE +/- 0.09, N = 3SE +/- 0.02, N = 323.4814.1313.0812.9612.6012.4511.0710.9410.8410.8110.3410.2610.1010.1010.1010.0610.0110.0010.00MIN: 10.06 / MAX: 112.91MIN: 10.63 / MAX: 167.28MIN: 10.11 / MAX: 444.45MIN: 10.23 / MAX: 424.46MIN: 9.82 / MAX: 418.4MIN: 11.55 / MAX: 14.48MIN: 10.1 / MAX: 13.23MIN: 9.95 / MAX: 12.7MIN: 9.93 / MAX: 12.81MIN: 9.95 / MAX: 12.78MIN: 10.14 / MAX: 11.37MIN: 10.09 / MAX: 11.22MIN: 9.97 / MAX: 11.42MIN: 9.84 / MAX: 11.72MIN: 9.86 / MAX: 11.08MIN: 9.95 / MAX: 11.04MIN: 9.88 / MAX: 11.4MIN: 9.91 / MAX: 11.15MIN: 9.92 / MAX: 12.351. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: resnet503070nv 40904090RTX 3070 Ti4090 rep3090 rep3090612182430SE +/- 0.24, N = 1523.5413.4612.9812.5211.2410.0410.03MIN: 10.3 / MAX: 149.49MIN: 10.6 / MAX: 340.67MIN: 10.26 / MAX: 145.62MIN: 9.95 / MAX: 459.05MIN: 10.22 / MAX: 29.96MIN: 9.94 / MAX: 10.91MIN: 9.88 / MAX: 10.861. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: blazeface3070RTX 3070 Ti4090 rep4080 xxx4080 rep4080 zzz4080ic3090 rep3090gdbafnv 409040900.67051.3412.01152.6823.3525SE +/- 0.14, N = 15SE +/- 0.01, N = 3SE +/- 0.00, N = 32.981.601.451.421.421.401.401.401.391.381.381.381.381.381.381.371.331.27MIN: 1.29 / MAX: 144.96MIN: 0.95 / MAX: 433.24MIN: 1.38 / MAX: 2.96MIN: 1.36 / MAX: 2.02MIN: 1.35 / MAX: 1.88MIN: 1.34 / MAX: 2.1MIN: 1.34 / MAX: 2.15MIN: 1.34 / MAX: 2MIN: 1.36 / MAX: 1.53MIN: 1.36 / MAX: 1.71MIN: 1.35 / MAX: 2.23MIN: 1.36 / MAX: 1.62MIN: 1.35 / MAX: 2.05MIN: 1.35 / MAX: 1.67MIN: 1.35 / MAX: 2.06MIN: 1.34 / MAX: 2.11MIN: 1.27 / MAX: 1.77MIN: 1.21 / MAX: 1.951. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: resnet50307040904090 repnv 4090RTX 3070 Ti4080 xxxi4080 zzz40804080 repgfc3090 rep3090b612182430SE +/- 0.27, N = 1523.1114.5813.5713.1312.3511.5011.1511.0910.9510.8410.4310.2510.0310.019.979.87MIN: 10.22 / MAX: 140.41MIN: 10.67 / MAX: 324.82MIN: 10.45 / MAX: 199.55MIN: 10.56 / MAX: 323.44MIN: 9.83 / MAX: 424.28MIN: 10.5 / MAX: 13.47MIN: 10.31 / MAX: 12.97MIN: 10.18 / MAX: 13.12MIN: 9.91 / MAX: 17.11MIN: 9.93 / MAX: 12.83MIN: 10.19 / MAX: 11.32MIN: 10.05 / MAX: 11.08MIN: 9.93 / MAX: 10.96MIN: 9.91 / MAX: 10.74MIN: 9.86 / MAX: 10.84MIN: 9.79 / MAX: 10.731. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: resnet183070nv 40904090RTX 3070 Ti4090 repi4080 xxx4080 rep40804080 zzzg3090 rep30903691215SE +/- 0.23, N = 1512.148.166.586.576.055.885.675.615.615.605.285.225.19MIN: 5.28 / MAX: 151.53MIN: 5.39 / MAX: 397.44MIN: 6.04 / MAX: 7.81MIN: 4.91 / MAX: 391.33MIN: 5.53 / MAX: 7.66MIN: 5.36 / MAX: 8.2MIN: 5.1 / MAX: 8.06MIN: 5.07 / MAX: 7.08MIN: 5.09 / MAX: 7.91MIN: 5.09 / MAX: 7.51MIN: 5.16 / MAX: 6.09MIN: 5.13 / MAX: 6.1MIN: 5.09 / MAX: 61. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: FastestDet3070nv 4090i4080 xxxRTX 3070 Ti4080f4090 rep4080 zzz3090 rep4080 repgb30904090c246810SE +/- 0.15, N = 158.635.924.434.314.264.204.204.164.124.104.094.064.064.043.943.69MIN: 4.27 / MAX: 144.3MIN: 4.25 / MAX: 103.26MIN: 4.28 / MAX: 5.01MIN: 4.14 / MAX: 6.11MIN: 2.71 / MAX: 347.03MIN: 4.04 / MAX: 5.82MIN: 4.15 / MAX: 4.92MIN: 4.03 / MAX: 4.73MIN: 3.97 / MAX: 6.99MIN: 4.06 / MAX: 4.21MIN: 3.92 / MAX: 5.5MIN: 4.01 / MAX: 4.78MIN: 4.03 / MAX: 4.3MIN: 4 / MAX: 4.15MIN: 3.8 / MAX: 5.41MIN: 3.66 / MAX: 3.921. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: alexnet3070RTX 3070 Ti4090 repnv 40904090i4080 rep4080 zzz4080 xxx4080g3090 rep30903691215SE +/- 0.22, N = 1510.085.535.335.145.145.104.724.684.674.614.354.314.31MIN: 4.36 / MAX: 225.66MIN: 4.22 / MAX: 362.62MIN: 4.83 / MAX: 6.6MIN: 4.65 / MAX: 6.81MIN: 4.76 / MAX: 6.16MIN: 4.75 / MAX: 6.12MIN: 4.25 / MAX: 7.3MIN: 4.26 / MAX: 6.8MIN: 4.27 / MAX: 6.36MIN: 4.24 / MAX: 7.25MIN: 4.28 / MAX: 5.1MIN: 4.26 / MAX: 5.07MIN: 4.25 / MAX: 4.941. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: googlenet30704090 repi4090RTX 3070 Tinv 40904080 rep40804080 zzz4080 xxxgacf3090 rep3090db48121620SE +/- 0.22, N = 15SE +/- 0.11, N = 3SE +/- 0.01, N = 318.2510.8610.3010.279.698.938.528.428.408.387.987.947.937.927.907.867.857.82MIN: 7.5 / MAX: 267.89MIN: 8.12 / MAX: 189.87MIN: 8.19 / MAX: 349.57MIN: 7.95 / MAX: 115.68MIN: 7.29 / MAX: 407.61MIN: 8.27 / MAX: 10.68MIN: 7.84 / MAX: 10.21MIN: 7.75 / MAX: 9.96MIN: 7.72 / MAX: 10.5MIN: 7.72 / MAX: 10.05MIN: 7.86 / MAX: 8.78MIN: 7.71 / MAX: 8.73MIN: 7.82 / MAX: 8.91MIN: 7.8 / MAX: 8.96MIN: 7.79 / MAX: 8.74MIN: 7.76 / MAX: 8.74MIN: 7.71 / MAX: 8.83MIN: 7.73 / MAX: 8.651. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: mnasnet3070nv 4090i4090 rep4090fRTX 3070 Ti4080 rep40804080 xxx4080 zzzgcb3090 rep3090246810SE +/- 0.04, N = 156.874.613.393.233.193.123.113.093.083.073.063.052.972.972.962.95MIN: 2.93 / MAX: 216.41MIN: 2.78 / MAX: 222.99MIN: 3.26 / MAX: 4.86MIN: 3.1 / MAX: 3.75MIN: 3.06 / MAX: 3.75MIN: 3.08 / MAX: 3.86MIN: 2.8 / MAX: 4.98MIN: 2.95 / MAX: 4.52MIN: 2.94 / MAX: 4.52MIN: 2.94 / MAX: 3.6MIN: 2.93 / MAX: 3.64MIN: 3.01 / MAX: 3.88MIN: 2.94 / MAX: 3.43MIN: 2.94 / MAX: 3.43MIN: 2.94 / MAX: 3.38MIN: 2.92 / MAX: 3.291. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: vision_transformer307040904090 repnv 4090iRTX 3070 Ti4080 xxx4080 rep40804080 zzzfg3090 rep3090bc1632486480SE +/- 0.13, N = 1573.5139.0138.7338.5838.3338.2935.4034.2934.1334.0533.3632.6832.1131.8631.7131.66MIN: 39.27 / MAX: 288.2MIN: 33.91 / MAX: 411.66MIN: 33.81 / MAX: 362.17MIN: 33.77 / MAX: 476.18MIN: 34.14 / MAX: 246.43MIN: 32.31 / MAX: 557.38MIN: 33.93 / MAX: 39.3MIN: 33.11 / MAX: 40.12MIN: 32.98 / MAX: 36.11MIN: 32.83 / MAX: 38.57MIN: 32.83 / MAX: 76.21MIN: 32.02 / MAX: 87.72MIN: 31.94 / MAX: 33.01MIN: 31.58 / MAX: 35.84MIN: 31.56 / MAX: 33.03MIN: 31.52 / MAX: 32.141. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: yolov4-tiny3070nv 409040904090 repRTX 3070 Ti4080i4080 xxx4080 zzz4080 repg30903090 rep714212835SE +/- 0.28, N = 1529.8017.3015.5515.4515.2113.8613.7713.6213.6113.5513.1413.1012.86MIN: 12.85 / MAX: 216.34MIN: 14.66 / MAX: 441.3MIN: 13.11 / MAX: 307.2MIN: 12.65 / MAX: 445.76MIN: 12.34 / MAX: 380.51MIN: 13.04 / MAX: 15.04MIN: 12.96 / MAX: 14.66MIN: 12.71 / MAX: 15.65MIN: 12.67 / MAX: 19.72MIN: 12.72 / MAX: 15.51MIN: 13 / MAX: 14.02MIN: 13.01 / MAX: 14.17MIN: 12.76 / MAX: 13.731. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: mobilenet3070nv 4090RTX 3070 Ti4090 rep40903090 rep3090510152025SE +/- 0.24, N = 1518.5410.159.988.838.468.058.01MIN: 8.01 / MAX: 164.45MIN: 8.08 / MAX: 193.04MIN: 7.79 / MAX: 434.9MIN: 8.29 / MAX: 10.15MIN: 8.12 / MAX: 10.14MIN: 7.98 / MAX: 8.94MIN: 7.96 / MAX: 8.471. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: yolov4-tiny3070nv 40904090 rep4090RTX 3070 Tiif4080 xxx40804080 rep4080 zzzg3090c3090 repb714212835SE +/- 0.23, N = 1529.4916.6116.3915.4415.4415.4314.3413.9513.7913.6813.6213.3512.8812.8612.8412.77MIN: 13.03 / MAX: 182.99MIN: 12.32 / MAX: 375.99MIN: 12.97 / MAX: 369.64MIN: 12.92 / MAX: 211.43MIN: 12.61 / MAX: 387.62MIN: 13.1 / MAX: 210.2MIN: 14.23 / MAX: 15.12MIN: 13.03 / MAX: 15.9MIN: 12.79 / MAX: 15.92MIN: 12.77 / MAX: 15.57MIN: 12.75 / MAX: 15.79MIN: 12.87 / MAX: 58.52MIN: 12.75 / MAX: 13.79MIN: 12.76 / MAX: 13.98MIN: 12.76 / MAX: 13.7MIN: 12.69 / MAX: 13.711. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v2-v2 - Model: mobilenet-v230704090nv 4090RTX 3070 Ti4080 xxx4090 rep4080 rep4080i4080 zzz3090 repg3090fcb246810SE +/- 0.18, N = 157.244.994.453.663.403.313.303.293.283.233.173.173.163.163.153.15MIN: 3.04 / MAX: 261.68MIN: 3.1 / MAX: 201.8MIN: 2.65 / MAX: 216.76MIN: 3.01 / MAX: 437.59MIN: 3.23 / MAX: 4.8MIN: 3.12 / MAX: 4.6MIN: 3.12 / MAX: 4.7MIN: 3.12 / MAX: 4.64MIN: 3.09 / MAX: 5.28MIN: 3.06 / MAX: 4.66MIN: 3.12 / MAX: 4.03MIN: 3.13 / MAX: 3.58MIN: 3.11 / MAX: 3.51MIN: 3.1 / MAX: 3.71MIN: 3.11 / MAX: 3.85MIN: 3.11 / MAX: 3.881. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: resnet1830704090RTX 3070 Ti4090 repnv 40903090 rep3691215SE +/- 0.05, N = 312.136.965.945.875.585.27MIN: 5.32 / MAX: 123.4MIN: 5.3 / MAX: 242.18MIN: 5.32 / MAX: 8.32MIN: 5.41 / MAX: 7.58MIN: 5.09 / MAX: 6.98MIN: 5.15 / MAX: 6.111. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: mobilenet-v33070nv 409040904090 repRTX 3070 Ti3090 rep246810SE +/- 0.04, N = 37.344.963.363.353.243.19MIN: 3.09 / MAX: 155.33MIN: 3.14 / MAX: 189.43MIN: 3.22 / MAX: 4.62MIN: 3.22 / MAX: 3.99MIN: 3.05 / MAX: 5.14MIN: 3.13 / MAX: 3.611. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: squeezenet_ssd3070iRTX 3070 Ti40904090 repnv 40904080 xxx40804080 rep4080 zzzgbf3090 repc309048121620SE +/- 0.19, N = 1516.158.338.297.937.817.727.707.667.637.517.317.147.097.087.077.04MIN: 7.25 / MAX: 210.69MIN: 6.32 / MAX: 222.03MIN: 6.37 / MAX: 448.22MIN: 7.31 / MAX: 9.45MIN: 7.24 / MAX: 9.04MIN: 7.12 / MAX: 23.25MIN: 7.11 / MAX: 9.19MIN: 7.02 / MAX: 9.08MIN: 7.02 / MAX: 9.71MIN: 6.94 / MAX: 9.51MIN: 6.96 / MAX: 30.1MIN: 7.06 / MAX: 7.95MIN: 6.98 / MAX: 8.01MIN: 7.01 / MAX: 7.93MIN: 7.01 / MAX: 7.75MIN: 6.97 / MAX: 7.761. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: alexnet30704090 repnv 40904090RTX 3070 Ti4080 xxxig4080 rep40804080 zzzbf3090 rep3090c3691215SE +/- 0.23, N = 159.866.586.116.115.675.214.994.864.694.694.654.424.354.304.304.30MIN: 4.25 / MAX: 157.02MIN: 4.61 / MAX: 91.07MIN: 4.83 / MAX: 124.76MIN: 4.73 / MAX: 81.72MIN: 4.21 / MAX: 365.75MIN: 4.79 / MAX: 6.66MIN: 4.59 / MAX: 6.56MIN: 4.8 / MAX: 6.37MIN: 4.26 / MAX: 7.17MIN: 4.26 / MAX: 6.15MIN: 4.26 / MAX: 5.97MIN: 4.32 / MAX: 5.1MIN: 4.27 / MAX: 5.16MIN: 4.25 / MAX: 4.7MIN: 4.25 / MAX: 5.08MIN: 4.26 / MAX: 5.161. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: yolov4-tiny3070nv 4090RTX 3070 Ti4090 rep4090i4080 zzz40804080 xxx4080 repgfbc30903090 rep714212835SE +/- 0.14, N = 1529.3415.6215.4215.3415.3015.1113.8313.8113.6913.5513.0813.0712.9812.8912.8712.82MIN: 12.17 / MAX: 245.34MIN: 12.99 / MAX: 184MIN: 12.21 / MAX: 414.81MIN: 12.94 / MAX: 157.95MIN: 12.87 / MAX: 144.73MIN: 12.93 / MAX: 151.45MIN: 12.89 / MAX: 15.4MIN: 12.84 / MAX: 15.1MIN: 12.73 / MAX: 15.68MIN: 12.75 / MAX: 14.74MIN: 12.96 / MAX: 13.83MIN: 12.95 / MAX: 14.55MIN: 12.73 / MAX: 35.55MIN: 12.84 / MAX: 13.19MIN: 12.75 / MAX: 13.58MIN: 12.72 / MAX: 13.481. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: mobilenet30704090iRTX 3070 Tinv 40904080g4080 zzz4080 xxx4080 rep4090 rep30903090 rep510152025SE +/- 0.27, N = 1518.3910.1810.029.629.418.848.508.468.468.408.378.118.04MIN: 7.92 / MAX: 173.39MIN: 8.18 / MAX: 235.56MIN: 8.07 / MAX: 266.25MIN: 7.71 / MAX: 449.11MIN: 8.98 / MAX: 11.38MIN: 8.31 / MAX: 10.98MIN: 8.42 / MAX: 9.29MIN: 7.97 / MAX: 10.56MIN: 7.95 / MAX: 10.34MIN: 7.93 / MAX: 15.25MIN: 7.98 / MAX: 10.71MIN: 8.02 / MAX: 14.2MIN: 7.96 / MAX: 9.011. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v2-v2 - Model: mobilenet-v23070RTX 3070 Ti4090 rep40904080 repnv 40904080 xxx3090 rep30904080 zzz246810SE +/- 0.53, N = 37.223.913.443.363.303.293.203.163.163.16MIN: 3.17 / MAX: 69.66MIN: 3.04 / MAX: 394.66MIN: 3.27 / MAX: 4.93MIN: 3.21 / MAX: 4.78MIN: 3.11 / MAX: 4.01MIN: 3.13 / MAX: 4.29MIN: 3.05 / MAX: 4.67MIN: 3.09 / MAX: 4.06MIN: 3.11 / MAX: 3.95MIN: 3.01 / MAX: 5.171. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: yolov4-tiny3070nv 409040904090 repRTX 3070 Ti3090 rep714212835SE +/- 0.94, N = 329.3816.3016.0515.4114.6412.92MIN: 12.95 / MAX: 201.31MIN: 14.11 / MAX: 184.46MIN: 12.93 / MAX: 474.03MIN: 12.75 / MAX: 226.87MIN: 12.77 / MAX: 383.28MIN: 12.79 / MAX: 18.51. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: regnety_400m30704090 rep40904080 repRTX 3070 Ti4080 xxx4080 zzznv 409030903090 rep48121620SE +/- 0.29, N = 318.2510.2310.098.728.428.388.348.258.228.03MIN: 7.8 / MAX: 238.29MIN: 8.22 / MAX: 197.1MIN: 8.01 / MAX: 418.58MIN: 8.32 / MAX: 10.48MIN: 7.66 / MAX: 10.74MIN: 8.04 / MAX: 9.63MIN: 8.03 / MAX: 10.23MIN: 7.87 / MAX: 10.07MIN: 8.14 / MAX: 8.67MIN: 7.97 / MAX: 8.651. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: vgg1630704090 repRTX 3070 Ti4090nv 40903090 rep1224364860SE +/- 0.28, N = 353.4829.8527.8627.3227.2523.72MIN: 25.52 / MAX: 296.52MIN: 24.25 / MAX: 400.86MIN: 24.17 / MAX: 416.36MIN: 24.36 / MAX: 262.38MIN: 24.12 / MAX: 252.53MIN: 23.56 / MAX: 24.591. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: squeezenet_ssd3070nv 4090RTX 3070 Ti4090 rep40904080 rep4080 zzz4080 xxx4080i3090gafdb3090 repce48121620SE +/- 0.26, N = 15SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 315.829.118.478.227.837.647.637.627.587.467.167.107.097.087.087.077.067.067.05MIN: 6.99 / MAX: 82.57MIN: 6.35 / MAX: 130.38MIN: 6.29 / MAX: 533.92MIN: 7.56 / MAX: 9.8MIN: 7.21 / MAX: 9.32MIN: 7.05 / MAX: 9.12MIN: 7 / MAX: 9.17MIN: 7.01 / MAX: 9.28MIN: 6.98 / MAX: 9.05MIN: 6.9 / MAX: 8.9MIN: 7.05 / MAX: 13.55MIN: 6.99 / MAX: 8.59MIN: 6.98 / MAX: 7.95MIN: 6.98 / MAX: 8.07MIN: 6.97 / MAX: 7.99MIN: 7 / MAX: 8.07MIN: 7 / MAX: 7.82MIN: 7 / MAX: 8.03MIN: 6.95 / MAX: 81. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: regnety_400m30704090nv 4090RTX 3070 Ti4090 rep40804080 xxx4080 zzzig3090 rep4080 rep309048121620SE +/- 0.21, N = 1517.8810.1010.099.198.648.618.528.498.468.388.348.247.99MIN: 7.38 / MAX: 190.77MIN: 7.93 / MAX: 156.75MIN: 7.84 / MAX: 366.66MIN: 7.44 / MAX: 524.66MIN: 8.3 / MAX: 10.51MIN: 8.21 / MAX: 10.07MIN: 8.13 / MAX: 9.73MIN: 8.08 / MAX: 9.72MIN: 8.08 / MAX: 10.33MIN: 8.05 / MAX: 27.34MIN: 8.26 / MAX: 9.09MIN: 7.91 / MAX: 9.53MIN: 7.92 / MAX: 8.781. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: alexnet3070RTX 3070 Tii4090 repnv 4090409040804080 zzz4080 xxx4080 repafgb3090 repc3090d3691215SE +/- 0.21, N = 14SE +/- 0.11, N = 3SE +/- 0.01, N = 39.625.495.305.235.185.144.754.694.684.674.414.364.324.324.314.314.304.30MIN: 4.31 / MAX: 147.6MIN: 4.26 / MAX: 363.39MIN: 4.92 / MAX: 7.18MIN: 4.78 / MAX: 7.33MIN: 4.75 / MAX: 7.12MIN: 4.73 / MAX: 6.32MIN: 4.31 / MAX: 13.88MIN: 4.29 / MAX: 5.78MIN: 4.28 / MAX: 6.37MIN: 4.27 / MAX: 5.88MIN: 4.24 / MAX: 5.16MIN: 4.29 / MAX: 5.7MIN: 4.25 / MAX: 5.17MIN: 4.26 / MAX: 5.15MIN: 4.26 / MAX: 5.18MIN: 4.26 / MAX: 4.98MIN: 4.25 / MAX: 4.83MIN: 4.23 / MAX: 5.321. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: yolov4-tiny30704090 repRTX 3070 Tinv 4090i40904080 zzz40804080 rep4080 xxxgf3090ebda3090 repc714212835SE +/- 0.18, N = 15SE +/- 0.04, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 328.5915.7215.5415.2615.1613.9713.8013.7913.6713.6513.6413.1712.8812.8712.8712.8512.8412.8312.81MIN: 12.87 / MAX: 325.37MIN: 13.2 / MAX: 301.81MIN: 12.15 / MAX: 492.01MIN: 12.87 / MAX: 132.82MIN: 12.86 / MAX: 248.64MIN: 13.11 / MAX: 16.15MIN: 12.76 / MAX: 15.76MIN: 12.75 / MAX: 19.63MIN: 12.71 / MAX: 14.88MIN: 12.71 / MAX: 14.99MIN: 13.04 / MAX: 76.32MIN: 13.03 / MAX: 14.1MIN: 12.76 / MAX: 13.67MIN: 12.68 / MAX: 13.84MIN: 12.76 / MAX: 13.73MIN: 12.72 / MAX: 13.93MIN: 12.69 / MAX: 15.33MIN: 12.74 / MAX: 13.59MIN: 12.73 / MAX: 13.081. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: regnety_400m30704090 rep4090RTX 3070 Ti4080 xxxf40804080 repb3090 rep4080 zzz3090igcnv 409048121620SE +/- 0.25, N = 1417.2310.349.608.898.588.508.458.448.278.258.108.017.997.997.987.73MIN: 7.8 / MAX: 193.14MIN: 8.21 / MAX: 214.16MIN: 7.66 / MAX: 210.23MIN: 7.74 / MAX: 476.28MIN: 8.23 / MAX: 10.39MIN: 8.04 / MAX: 30.12MIN: 8.05 / MAX: 10.3MIN: 8.04 / MAX: 10.17MIN: 8.22 / MAX: 9.01MIN: 8.12 / MAX: 14MIN: 7.77 / MAX: 15.42MIN: 7.93 / MAX: 8.35MIN: 7.62 / MAX: 9.27MIN: 7.91 / MAX: 8.8MIN: 7.93 / MAX: 8.65MIN: 7.43 / MAX: 9.411. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: regnety_400m3070nv 4090i4090 repRTX 3070 Ti4080 rep4080 xxx4080 zzzgc30903090 rep4080dba4090f48121620SE +/- 0.19, N = 15SE +/- 0.06, N = 3SE +/- 0.04, N = 318.0010.179.949.668.838.678.458.378.308.278.258.248.248.238.188.188.138.08MIN: 7.91 / MAX: 176.28MIN: 8.12 / MAX: 209.53MIN: 7.43 / MAX: 166.02MIN: 7.78 / MAX: 95.3MIN: 7.65 / MAX: 351.08MIN: 8.22 / MAX: 15.29MIN: 8.12 / MAX: 9.68MIN: 8.05 / MAX: 10.19MIN: 8.22 / MAX: 9.1MIN: 8.22 / MAX: 9.18MIN: 8.17 / MAX: 8.9MIN: 8.17 / MAX: 8.84MIN: 7.89 / MAX: 9.52MIN: 8.03 / MAX: 8.9MIN: 8.12 / MAX: 8.86MIN: 8.07 / MAX: 9.68MIN: 7.78 / MAX: 9.98MIN: 7.98 / MAX: 10.871. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: mobilenet3070iRTX 3070 Tig4090nv 40904090 repf4080 rep4080 xxx40804080 zzz30903090 repbc48121620SE +/- 0.23, N = 1517.8210.089.628.988.968.918.748.568.488.448.438.408.068.038.018.00MIN: 7.57 / MAX: 211.62MIN: 8.08 / MAX: 286.28MIN: 7.76 / MAX: 454.91MIN: 8.1 / MAX: 124.43MIN: 8.39 / MAX: 10.77MIN: 8.33 / MAX: 10.07MIN: 8.25 / MAX: 10.5MIN: 8.04 / MAX: 75.44MIN: 7.96 / MAX: 10.32MIN: 7.97 / MAX: 10.71MIN: 7.99 / MAX: 10.44MIN: 8.12 / MAX: 10.11MIN: 7.94 / MAX: 13.92MIN: 7.98 / MAX: 8.77MIN: 7.95 / MAX: 8.95MIN: 7.96 / MAX: 8.631. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: yolov4-tiny30704090nv 40904090 rep4080 zzzRTX 3070 Ti40804080 rep4080 xxx30903090 rep714212835SE +/- 0.19, N = 1528.7315.9515.5515.4015.2615.0013.9313.7313.5212.9712.90MIN: 12.83 / MAX: 264.49MIN: 13.38 / MAX: 245.18MIN: 12.87 / MAX: 342.3MIN: 13 / MAX: 245.79MIN: 14.19 / MAX: 17.06MIN: 12.75 / MAX: 401.37MIN: 13.08 / MAX: 15.68MIN: 12.78 / MAX: 20.99MIN: 12.72 / MAX: 21.19MIN: 12.83 / MAX: 13.8MIN: 12.77 / MAX: 13.921. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: regnety_400m30704090RTX 3070 Ti4090 repnv 409030903090 rep48121620SE +/- 0.20, N = 1518.2410.059.108.458.378.208.19MIN: 7.5 / MAX: 201.09MIN: 8.13 / MAX: 173.18MIN: 7.61 / MAX: 454.62MIN: 8.05 / MAX: 12.64MIN: 8.08 / MAX: 10.1MIN: 8.14 / MAX: 8.74MIN: 8.12 / MAX: 8.981. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: vision_transformer3070nv 40904090RTX 3070 Ti4090 repi40804080 rep4080 zzz4080 xxxfgd3090e3090 repabc1632486480SE +/- 0.16, N = 15SE +/- 0.21, N = 3SE +/- 0.07, N = 3SE +/- 0.09, N = 370.7639.0438.7638.0337.5936.4235.5635.0734.3234.1932.9232.4232.1231.9431.9331.9131.8831.8531.79MIN: 38.81 / MAX: 250.01MIN: 33.83 / MAX: 463.88MIN: 33.12 / MAX: 539.58MIN: 32.66 / MAX: 467.28MIN: 34.45 / MAX: 457.98MIN: 33.49 / MAX: 224.86MIN: 33.19 / MAX: 40.43MIN: 33.66 / MAX: 39.36MIN: 32.58 / MAX: 41.88MIN: 32.72 / MAX: 36.79MIN: 32.67 / MAX: 36.93MIN: 31.89 / MAX: 65.47MIN: 31.66 / MAX: 46.9MIN: 31.73 / MAX: 34.21MIN: 31.62 / MAX: 35.85MIN: 31.74 / MAX: 34.28MIN: 31.55 / MAX: 37.47MIN: 31.69 / MAX: 33.06MIN: 31.63 / MAX: 35.571. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: yolov4-tiny3070nv 40904090RTX 3070 Ti4090 rep4080 rep4080 xxx4080 zzz30903090 rep714212835SE +/- 0.81, N = 328.4117.6715.8514.5713.8813.7113.6313.4212.8212.81MIN: 12.49 / MAX: 151.04MIN: 14.92 / MAX: 343.93MIN: 13.26 / MAX: 253.23MIN: 12.33 / MAX: 312.42MIN: 13.09 / MAX: 14.77MIN: 12.78 / MAX: 15.62MIN: 12.77 / MAX: 16.93MIN: 12.65 / MAX: 16.19MIN: 12.72 / MAX: 13.66MIN: 12.7 / MAX: 13.691. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: vision_transformer3070nv 409040904090 repiRTX 3070 Ti40804080 xxx4080 zzz4080 repg30903090 rep1632486480SE +/- 0.20, N = 1571.0839.1839.0638.1738.0137.8834.9134.2334.1034.1033.3233.2232.09MIN: 38.84 / MAX: 374.68MIN: 33.74 / MAX: 520.24MIN: 34.16 / MAX: 481.28MIN: 32.97 / MAX: 462.63MIN: 32.96 / MAX: 388.09MIN: 32.46 / MAX: 518.57MIN: 33.72 / MAX: 36.82MIN: 33.08 / MAX: 37.43MIN: 32.32 / MAX: 38.54MIN: 32.43 / MAX: 38.75MIN: 31.83 / MAX: 104.12MIN: 33.04 / MAX: 36.99MIN: 31.84 / MAX: 32.771. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: resnet503070nv 4090RTX 3070 Ti4090 rep40904080 xxx4080 zzz4080 rep3090 rep3090510152025SE +/- 0.04, N = 322.1913.2912.8112.4711.7211.2611.1010.8610.0610.03MIN: 10.16 / MAX: 181.74MIN: 10.54 / MAX: 456.82MIN: 10.06 / MAX: 349.03MIN: 11.5 / MAX: 14.68MIN: 10.8 / MAX: 12.8MIN: 10.32 / MAX: 13.29MIN: 10.19 / MAX: 18.3MIN: 9.98 / MAX: 12.46MIN: 9.86 / MAX: 11.9MIN: 9.93 / MAX: 10.871. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: vision_transformer307040904090 repRTX 3070 Tinv 40903090 rep1632486480SE +/- 0.10, N = 370.5339.3538.6538.5037.1331.94MIN: 39.2 / MAX: 276.33MIN: 34.22 / MAX: 466.65MIN: 33.07 / MAX: 476.08MIN: 33.7 / MAX: 418.06MIN: 33.97 / MAX: 443.1MIN: 31.73 / MAX: 32.751. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: vision_transformer30704090 rep4090nv 4090RTX 3070 Ti40804080 zzz4080 rep4080 xxx3090 rep30901632486480SE +/- 0.11, N = 1570.2939.0338.6238.4638.2735.6035.3633.9333.9031.9731.94MIN: 39.39 / MAX: 250.19MIN: 33.61 / MAX: 343.67MIN: 33.33 / MAX: 465MIN: 32.39 / MAX: 435.46MIN: 32.29 / MAX: 507.7MIN: 34.13 / MAX: 38.49MIN: 33.87 / MAX: 42.41MIN: 32.77 / MAX: 36.2MIN: 32.72 / MAX: 37.77MIN: 31.71 / MAX: 33.78MIN: 31.72 / MAX: 34.341. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: efficientnet-b03070RTX 3070 Ti4090inv 409040804090 rep4080 xxx4080 zzz4080 rep3090 repg3090246810SE +/- 0.19, N = 158.414.604.364.194.124.064.044.024.013.983.873.843.83MIN: 3.76 / MAX: 67.73MIN: 3.79 / MAX: 336.2MIN: 4.14 / MAX: 5.24MIN: 4.01 / MAX: 5.09MIN: 3.86 / MAX: 5.39MIN: 3.85 / MAX: 4.97MIN: 3.85 / MAX: 4.9MIN: 3.8 / MAX: 5.14MIN: 3.79 / MAX: 5.39MIN: 3.77 / MAX: 5.44MIN: 3.81 / MAX: 4.62MIN: 3.78 / MAX: 4.57MIN: 3.78 / MAX: 4.41. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: squeezenet_ssd30704090 repRTX 3070 Tinv 409040903090 rep309048121620SE +/- 0.22, N = 1515.469.348.317.727.407.077.05MIN: 7.08 / MAX: 147.31MIN: 6.88 / MAX: 268.7MIN: 6.35 / MAX: 364.95MIN: 7.13 / MAX: 8.97MIN: 6.81 / MAX: 8.46MIN: 6.99 / MAX: 7.81MIN: 6.98 / MAX: 7.811. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: squeezenet_ssd30704090nv 40904090 repRTX 3070 Ti4080 xxx40804080 zzz4080 repig3090 rep309048121620SE +/- 0.25, N = 1415.409.289.219.168.287.677.667.627.557.217.147.097.04MIN: 6.64 / MAX: 132.68MIN: 6.92 / MAX: 355.6MIN: 6.83 / MAX: 203.62MIN: 6.73 / MAX: 423.75MIN: 6.38 / MAX: 381.81MIN: 7.04 / MAX: 9.1MIN: 7.09 / MAX: 8.97MIN: 7 / MAX: 9.93MIN: 6.99 / MAX: 9.08MIN: 6.73 / MAX: 8.82MIN: 7.03 / MAX: 7.99MIN: 7.01 / MAX: 7.97MIN: 6.96 / MAX: 7.71. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: vgg163070nv 40904090 repRTX 3070 Ti409030903090 rep1224364860SE +/- 0.26, N = 1551.2829.4029.3528.5327.3123.5023.47MIN: 24.83 / MAX: 242.12MIN: 26.17 / MAX: 411.51MIN: 24.55 / MAX: 485.35MIN: 24.21 / MAX: 515.3MIN: 24.27 / MAX: 230.86MIN: 23.26 / MAX: 24.34MIN: 23.25 / MAX: 24.241. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: regnety_400m30704090 repnv 40904090RTX 3070 Ti40804080 zzz4080 rep30904080 xxx3090 rep48121620SE +/- 0.24, N = 1517.6110.6910.039.879.058.678.378.358.338.258.07MIN: 7.85 / MAX: 165.34MIN: 8.17 / MAX: 339.6MIN: 7.81 / MAX: 171.2MIN: 7.81 / MAX: 243.06MIN: 7.52 / MAX: 417.33MIN: 8.3 / MAX: 14.66MIN: 8.04 / MAX: 10.13MIN: 8.05 / MAX: 9.76MIN: 8.25 / MAX: 9.32MIN: 7.93 / MAX: 9.88MIN: 7.99 / MAX: 8.881. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: resnet1830704090nv 4090RTX 3070 Ti40804090 rep4080 zzz4080 rep4080 xxx30903090 rep3691215SE +/- 0.20, N = 1511.307.527.386.405.925.905.745.675.655.215.20MIN: 5.3 / MAX: 181.7MIN: 5.45 / MAX: 290.49MIN: 5.15 / MAX: 138.85MIN: 5.1 / MAX: 457.07MIN: 5.37 / MAX: 8.24MIN: 5.43 / MAX: 7.49MIN: 5.18 / MAX: 8.08MIN: 5.19 / MAX: 7.38MIN: 5.18 / MAX: 6.76MIN: 5.09 / MAX: 6.13MIN: 5.1 / MAX: 6.091. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: googlenet3070i4090RTX 3070 Ti4090 repgnv 40904080 zzz4080 rep4080 xxx4080fc3090 repb309048121620SE +/- 0.24, N = 1517.0010.179.979.659.299.159.028.558.528.508.458.077.887.857.847.83MIN: 7.35 / MAX: 277.79MIN: 7.94 / MAX: 150.01MIN: 7.67 / MAX: 258.52MIN: 7.59 / MAX: 472.81MIN: 7.98 / MAX: 83.03MIN: 7.84 / MAX: 198.46MIN: 8.41 / MAX: 11.08MIN: 7.85 / MAX: 10.35MIN: 7.81 / MAX: 10.78MIN: 7.79 / MAX: 9.94MIN: 7.79 / MAX: 10.32MIN: 7.92 / MAX: 8.86MIN: 7.79 / MAX: 8.78MIN: 7.75 / MAX: 8.69MIN: 7.74 / MAX: 8.7MIN: 7.71 / MAX: 8.81. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: yolov4-tiny3070g4090 repnv 4090RTX 3070 Tii30904080 rep408040904080 zzz4080 xxxfdac3090 repb714212835SE +/- 0.18, N = 15SE +/- 0.05, N = 3SE +/- 0.11, N = 327.6617.2316.7915.4015.2014.6514.2614.0313.8513.6813.6313.6013.3212.9512.9012.8112.7712.74MIN: 12.74 / MAX: 294.9MIN: 12.99 / MAX: 196.66MIN: 14.1 / MAX: 273.41MIN: 12.35 / MAX: 321.43MIN: 12.69 / MAX: 431.37MIN: 12.44 / MAX: 202.68MIN: 14.17 / MAX: 14.53MIN: 13.15 / MAX: 15.97MIN: 12.84 / MAX: 16.75MIN: 12.83 / MAX: 14.63MIN: 12.77 / MAX: 15.36MIN: 12.8 / MAX: 16.23MIN: 12.95 / MAX: 35.49MIN: 12.75 / MAX: 18.88MIN: 12.69 / MAX: 15.88MIN: 12.74 / MAX: 13.2MIN: 12.7 / MAX: 13.02MIN: 12.66 / MAX: 13.281. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: squeezenet_ssd30704090 repnv 4090RTX 3070 Ti40903090 rep48121620SE +/- 0.14, N = 315.329.448.267.457.437.07MIN: 6.66 / MAX: 139.17MIN: 7.17 / MAX: 94.63MIN: 7.64 / MAX: 11.08MIN: 6.59 / MAX: 9.11MIN: 6.84 / MAX: 8.82MIN: 6.98 / MAX: 9.711. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: googlenet30704090 repRTX 3070 Ti409040804080 zzz4080 repnv 40904080 xxx3090 rep309048121620SE +/- 0.19, N = 1516.9710.399.909.058.798.558.428.358.267.867.84MIN: 7.44 / MAX: 229.93MIN: 7.87 / MAX: 391.66MIN: 7.76 / MAX: 396.66MIN: 8.26 / MAX: 13.34MIN: 8.08 / MAX: 10.27MIN: 7.86 / MAX: 10.08MIN: 7.77 / MAX: 10.52MIN: 7.7 / MAX: 10.46MIN: 7.62 / MAX: 10.47MIN: 7.75 / MAX: 8.71MIN: 7.74 / MAX: 8.721. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: vision_transformer307040904090 repnv 4090RTX 3070 Ti30903090 rep1530456075SE +/- 0.12, N = 1569.4838.8238.6938.5838.3232.1632.13MIN: 39.08 / MAX: 374.31MIN: 33.83 / MAX: 435.6MIN: 33.32 / MAX: 390.07MIN: 33.06 / MAX: 464.16MIN: 32.26 / MAX: 477.15MIN: 31.94 / MAX: 33.7MIN: 31.95 / MAX: 32.871. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: resnet50307040904090 repRTX 3070 Tii4080 repnv 409040804080 zzzf4080 xxxg3090acbd3090 rep510152025SE +/- 0.26, N = 15SE +/- 0.23, N = 3SE +/- 0.01, N = 321.5014.1013.7312.7312.0911.7611.4111.1611.1011.0510.8210.7210.3010.2010.1110.0110.009.95MIN: 10.24 / MAX: 116.85MIN: 10.27 / MAX: 287MIN: 10.4 / MAX: 137.78MIN: 10.18 / MAX: 541.92MIN: 11.16 / MAX: 13.48MIN: 10.68 / MAX: 44.94MIN: 10.57 / MAX: 12.22MIN: 10.29 / MAX: 15.03MIN: 10.2 / MAX: 13.06MIN: 10.14 / MAX: 162.88MIN: 9.9 / MAX: 12.26MIN: 10.1 / MAX: 108.3MIN: 9.82 / MAX: 17.56MIN: 9.84 / MAX: 12.48MIN: 9.95 / MAX: 16.18MIN: 9.85 / MAX: 11.06MIN: 9.86 / MAX: 11.02MIN: 9.85 / MAX: 10.721. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: resnet503070nv 4090RTX 3070 Ti40904090 rep3090 rep510152025SE +/- 0.30, N = 322.1513.2513.1513.0010.9610.27MIN: 10.11 / MAX: 123.04MIN: 10.61 / MAX: 154.12MIN: 10.26 / MAX: 349.93MIN: 10.34 / MAX: 397.57MIN: 10.09 / MAX: 12.99MIN: 10.12 / MAX: 11.191. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: vgg1630704090 rep4090RTX 3070 Tinv 40904080 xxx4080 zzz4080 rep30903090 rep1122334455SE +/- 0.54, N = 350.3230.7430.1627.9827.6125.4425.2625.0123.5823.40MIN: 25.92 / MAX: 281.06MIN: 25.36 / MAX: 428.68MIN: 24.66 / MAX: 332.49MIN: 24.35 / MAX: 423.63MIN: 24.67 / MAX: 401.29MIN: 24.27 / MAX: 27.68MIN: 24.29 / MAX: 27.75MIN: 23.88 / MAX: 26.66MIN: 23.35 / MAX: 24.43MIN: 23.2 / MAX: 24.071. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: resnet1830704090 repnv 4090RTX 3070 Tifi40904080 zzz40804080 xxx4080 repgcb3090 rep30903691215SE +/- 0.19, N = 1511.147.757.616.186.135.865.815.715.705.665.635.485.265.215.205.19MIN: 4.79 / MAX: 65.12MIN: 5.57 / MAX: 125.43MIN: 5.23 / MAX: 90.18MIN: 5.17 / MAX: 262.79MIN: 5.41 / MAX: 151.51MIN: 5.35 / MAX: 7.79MIN: 5.27 / MAX: 7.16MIN: 5.12 / MAX: 8.19MIN: 5.15 / MAX: 7.9MIN: 5.14 / MAX: 7.49MIN: 5.09 / MAX: 7.75MIN: 5.37 / MAX: 6.51MIN: 5.18 / MAX: 6.27MIN: 5.12 / MAX: 6.22MIN: 5.1 / MAX: 5.97MIN: 5.09 / MAX: 6.131. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: FastestDet3070i4080 zzz4090 rep4090RTX 3070 Tif4080 rep4080 xxx40803090dac3090 repegbnv 4090246810SE +/- 0.29, N = 15SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 38.415.144.794.594.394.264.244.214.204.204.114.114.104.094.084.084.074.073.93MIN: 2.89 / MAX: 487.78MIN: 3.7 / MAX: 81.79MIN: 4.64 / MAX: 6.21MIN: 2.62 / MAX: 232.18MIN: 4.25 / MAX: 5.86MIN: 2.5 / MAX: 396.93MIN: 3.88 / MAX: 24.21MIN: 4.04 / MAX: 4.97MIN: 4.03 / MAX: 6.49MIN: 4.02 / MAX: 4.97MIN: 4.07 / MAX: 4.29MIN: 4.01 / MAX: 9.72MIN: 4.06 / MAX: 4.81MIN: 4.05 / MAX: 5.5MIN: 4.04 / MAX: 4.35MIN: 4.03 / MAX: 5.29MIN: 4.02 / MAX: 4.82MIN: 4.04 / MAX: 4.53MIN: 3.8 / MAX: 5.41. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: shufflenet-v23070RTX 3070 Tinv 40904080 xxx4090 repi4080 zzz40904080 rep4080f30903090 repgdaebc246810SE +/- 0.20, N = 15SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 37.073.983.513.503.493.493.463.453.443.433.403.393.363.353.353.353.333.333.32MIN: 3.25 / MAX: 243.32MIN: 3.14 / MAX: 529.82MIN: 3.37 / MAX: 4MIN: 3.37 / MAX: 4.85MIN: 3.36 / MAX: 4.33MIN: 3.35 / MAX: 4.24MIN: 3.32 / MAX: 5.24MIN: 3.32 / MAX: 3.99MIN: 3.3 / MAX: 5.36MIN: 3.3 / MAX: 4.22MIN: 3.35 / MAX: 5.89MIN: 3.35 / MAX: 3.69MIN: 3.32 / MAX: 4.06MIN: 3.3 / MAX: 4.02MIN: 3.3 / MAX: 3.82MIN: 3.29 / MAX: 3.85MIN: 3.28 / MAX: 4.14MIN: 3.3 / MAX: 3.59MIN: 3.29 / MAX: 4.191. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: mobilenet3070RTX 3070 Ti4080 zzz40904090 repnv 409040804080 rep4080 xxx3090 rep309048121620SE +/- 0.25, N = 1517.099.529.199.169.028.938.438.388.318.058.03MIN: 7.89 / MAX: 121.53MIN: 7.97 / MAX: 420.29MIN: 8.51 / MAX: 11.04MIN: 8.5 / MAX: 10.51MIN: 8.42 / MAX: 11.17MIN: 8.33 / MAX: 11.07MIN: 8.03 / MAX: 9.64MIN: 7.94 / MAX: 10.07MIN: 7.85 / MAX: 10.21MIN: 7.96 / MAX: 9.04MIN: 7.96 / MAX: 8.831. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: vgg1630704090 repRTX 3070 Tinv 409040904080 zzz40804080 rep4080 xxx30903090 rep1122334455SE +/- 0.28, N = 1549.7029.1728.5328.1427.4426.0925.6725.5625.3323.5123.38MIN: 25.55 / MAX: 421.44MIN: 24.61 / MAX: 264.85MIN: 23.95 / MAX: 473.83MIN: 24.24 / MAX: 221.5MIN: 24.06 / MAX: 264.59MIN: 24.58 / MAX: 30.18MIN: 24.46 / MAX: 27.34MIN: 24.24 / MAX: 27.92MIN: 24.26 / MAX: 34.98MIN: 23.27 / MAX: 24.38MIN: 23.19 / MAX: 24.271. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: mobilenet3070nv 40904090RTX 3070 Ti4090 rep3090 rep48121620SE +/- 0.13, N = 317.0610.6410.5610.028.228.03MIN: 8 / MAX: 101.45MIN: 8.4 / MAX: 127.99MIN: 8.32 / MAX: 239.95MIN: 7.8 / MAX: 372.36MIN: 7.75 / MAX: 9.41MIN: 7.97 / MAX: 8.911. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3 - Model: vgg163070i4090RTX 3070 Ti4090 repnv 40904080 zzz40804080 xxxg4080 repfc3090 rep3090b1122334455SE +/- 0.24, N = 1549.7529.1228.5528.4027.2527.0425.4525.1025.0024.9224.9124.4523.9923.5423.5023.50MIN: 25.45 / MAX: 273.86MIN: 26.33 / MAX: 310.23MIN: 24.05 / MAX: 201.8MIN: 24.12 / MAX: 509.06MIN: 24.14 / MAX: 379.93MIN: 24.33 / MAX: 215.56MIN: 24.22 / MAX: 27.73MIN: 24.12 / MAX: 27.57MIN: 23.91 / MAX: 27.99MIN: 24.58 / MAX: 31.89MIN: 23.8 / MAX: 26.87MIN: 24.26 / MAX: 25.26MIN: 23.72 / MAX: 24.98MIN: 23.33 / MAX: 24.41MIN: 23.17 / MAX: 24.44MIN: 23.3 / MAX: 24.411. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: regnety_400m30704090RTX 3070 Ti4090 repnv 40903090 rep48121620SE +/- 0.54, N = 317.0210.119.148.708.348.06MIN: 7.65 / MAX: 216.63MIN: 8.03 / MAX: 259.38MIN: 8.14 / MAX: 400.02MIN: 8.29 / MAX: 12.6MIN: 8.01 / MAX: 12.36MIN: 7.98 / MAX: 8.61. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3 - Model: mobilenet-v33070iRTX 3070 Tinv 40904090 rep40904080 zzzc3090eda3090 repf246810SE +/- 0.20, N = 14SE +/- 0.02, N = 3SE +/- 0.00, N = 2SE +/- 0.02, N = 36.604.873.643.363.303.283.283.203.193.183.173.173.163.14MIN: 2.98 / MAX: 166.19MIN: 3.14 / MAX: 278.98MIN: 2.87 / MAX: 429.02MIN: 3.21 / MAX: 4.3MIN: 3.15 / MAX: 3.92MIN: 3.15 / MAX: 3.9MIN: 3.13 / MAX: 4.65MIN: 3.16 / MAX: 3.68MIN: 3.14 / MAX: 3.48MIN: 3.11 / MAX: 3.78MIN: 3.1 / MAX: 3.83MIN: 3.11 / MAX: 3.73MIN: 3.11 / MAX: 3.62MIN: 3.09 / MAX: 3.541. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3 - Model: mobilenet-v33070nv 4090RTX 3070 Ti40904090 rep4080 rep3090 rep4080 xxx4080 zzz246810SE +/- 0.53, N = 36.434.973.703.333.303.283.173.083.06MIN: 2.85 / MAX: 164.91MIN: 3.15 / MAX: 291.01MIN: 2.98 / MAX: 261.6MIN: 3.2 / MAX: 4.4MIN: 3.15 / MAX: 3.91MIN: 3.13 / MAX: 4.78MIN: 3.12 / MAX: 3.75MIN: 2.97 / MAX: 3.67MIN: 2.94 / MAX: 3.941. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3 - Model: mobilenet-v33070RTX 3070 Ti40904090 repnv 40904080 xxxi4080 zzz4080 rep3090g246810SE +/- 0.18, N = 156.563.623.343.333.263.263.263.243.243.163.16MIN: 3.07 / MAX: 110.87MIN: 3 / MAX: 469.9MIN: 3.21 / MAX: 4.31MIN: 3.19 / MAX: 4.79MIN: 3.13 / MAX: 3.96MIN: 3.13 / MAX: 4.08MIN: 3.11 / MAX: 4.7MIN: 3.1 / MAX: 3.88MIN: 3.11 / MAX: 4.37MIN: 3.11 / MAX: 3.77MIN: 3.12 / MAX: 3.581. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: mnasnet30704090 rep4090RTX 3070 Ti40804080 zzznv 40904080 rep4080 xxx3090 rep3090246810SE +/- 0.04, N = 156.064.994.933.103.103.083.073.062.982.972.94MIN: 2.96 / MAX: 42.7MIN: 3.02 / MAX: 235.56MIN: 2.97 / MAX: 124.96MIN: 2.61 / MAX: 4.75MIN: 2.95 / MAX: 4.05MIN: 2.95 / MAX: 3.88MIN: 2.93 / MAX: 4.52MIN: 2.93 / MAX: 5.02MIN: 2.86 / MAX: 4.47MIN: 2.94 / MAX: 3.45MIN: 2.9 / MAX: 3.341. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: vgg163070nv 4090RTX 3070 Ti4090i4090 rep4080 zzz4080 rep4080 xxx4080fgedb3090ca3090 rep1122334455SE +/- 0.24, N = 15SE +/- 0.14, N = 3SE +/- 0.10, N = 3SE +/- 0.04, N = 348.2929.2929.0628.8227.8327.0425.8225.0425.0125.0024.5524.2023.6023.5623.5623.5523.5423.5123.48MIN: 24.97 / MAX: 183.12MIN: 24.63 / MAX: 296.95MIN: 24.11 / MAX: 541.55MIN: 24.35 / MAX: 214.1MIN: 24.98 / MAX: 262.23MIN: 24.22 / MAX: 296.13MIN: 24.35 / MAX: 62.94MIN: 24.06 / MAX: 27.35MIN: 23.8 / MAX: 26.41MIN: 23.93 / MAX: 26.69MIN: 23.62 / MAX: 97.69MIN: 23.56 / MAX: 58.31MIN: 23.17 / MAX: 24.71MIN: 23.24 / MAX: 24.78MIN: 23.34 / MAX: 24.72MIN: 23.3 / MAX: 24.45MIN: 23.33 / MAX: 24.61MIN: 23.29 / MAX: 24.68MIN: 23.24 / MAX: 29.211. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: mobilenet3070nv 4090RTX 3070 Ti4090 repi4080 xxx4090f4080 rep40804080 zzzg3090 rep3090bc48121620SE +/- 0.26, N = 1516.3410.549.629.549.058.888.818.658.468.438.388.208.018.008.007.95MIN: 8.13 / MAX: 80.69MIN: 8.41 / MAX: 134.08MIN: 7.76 / MAX: 502.83MIN: 8.94 / MAX: 10.54MIN: 8.48 / MAX: 11.28MIN: 8.31 / MAX: 10.01MIN: 8.32 / MAX: 10.7MIN: 8.55 / MAX: 9.53MIN: 7.99 / MAX: 10.62MIN: 7.99 / MAX: 10.66MIN: 7.95 / MAX: 10.41MIN: 8.12 / MAX: 9.4MIN: 7.95 / MAX: 8.35MIN: 7.94 / MAX: 8.78MIN: 7.95 / MAX: 8.99MIN: 7.89 / MAX: 8.791. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: vision_transformer3070nv 40904090 rep4090RTX 3070 Ti4080 zzz4080 rep4080 xxx30903090 rep1530456075SE +/- 0.11, N = 365.4138.9538.7938.7638.0434.4734.2234.1432.1031.85MIN: 39.08 / MAX: 230.59MIN: 34.04 / MAX: 486.96MIN: 34.02 / MAX: 460.15MIN: 33.38 / MAX: 423.24MIN: 33.11 / MAX: 346.94MIN: 33.05 / MAX: 39.69MIN: 33.01 / MAX: 37.09MIN: 32.5 / MAX: 37.13MIN: 31.9 / MAX: 33.03MIN: 31.67 / MAX: 35.741. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: mnasnet307040904090 repRTX 3070 Tinv 40904080 rep4080 xxx30903090 rep4080 zzz246810SE +/- 0.13, N = 36.075.195.113.243.123.093.002.982.972.96MIN: 2.94 / MAX: 129.1MIN: 3.04 / MAX: 436.91MIN: 2.96 / MAX: 247.47MIN: 2.9 / MAX: 5.34MIN: 2.98 / MAX: 3.71MIN: 2.96 / MAX: 4.98MIN: 2.88 / MAX: 4.37MIN: 2.95 / MAX: 3.9MIN: 2.93 / MAX: 3.3MIN: 2.85 / MAX: 3.821. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: mobilenet3070nv 40904090 repRTX 3070 Ti40904080 rep4080 xxx4080 zzz30903090 rep48121620SE +/- 0.14, N = 316.5212.1210.6110.039.048.458.348.258.078.06MIN: 7.9 / MAX: 82.53MIN: 9.16 / MAX: 505.01MIN: 8.34 / MAX: 225.97MIN: 7.86 / MAX: 346.64MIN: 8.49 / MAX: 10.96MIN: 8.01 / MAX: 10.86MIN: 7.89 / MAX: 9.42MIN: 7.78 / MAX: 9.61MIN: 8.01 / MAX: 8.62MIN: 8 / MAX: 8.961. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: yolov4-tiny30704090 rep4090nv 4090RTX 3070 Ti3090 rep3090612182430SE +/- 0.25, N = 1526.3316.6015.6915.6715.5612.8912.86MIN: 12.62 / MAX: 127.32MIN: 12.98 / MAX: 103.04MIN: 13.13 / MAX: 187.93MIN: 12.91 / MAX: 334.44MIN: 12.24 / MAX: 459.8MIN: 12.79 / MAX: 13.77MIN: 12.74 / MAX: 13.681. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: efficientnet-b030704090 repnv 4090RTX 3070 Tig4080 xxxi409040804080 rep4080 zzz3090 rep3090fbc246810SE +/- 0.18, N = 157.816.285.264.734.634.224.214.184.054.043.953.853.853.853.853.82MIN: 3.73 / MAX: 159.47MIN: 3.91 / MAX: 337.73MIN: 3.48 / MAX: 250.88MIN: 3.79 / MAX: 418.72MIN: 3.8 / MAX: 159.43MIN: 4 / MAX: 5.58MIN: 3.96 / MAX: 4.94MIN: 4 / MAX: 5.25MIN: 3.83 / MAX: 5MIN: 3.81 / MAX: 5.08MIN: 3.76 / MAX: 4.84MIN: 3.81 / MAX: 4.62MIN: 3.8 / MAX: 4.43MIN: 3.8 / MAX: 4.6MIN: 3.82 / MAX: 4.48MIN: 3.78 / MAX: 4.531. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: shufflenet-v23070iRTX 3070 Tig4090f4090 repnv 40904080 xxx4080 rep4080 zzz40803090 repdc3090ba246810SE +/- 0.19, N = 15SE +/- 0.01, N = 3SE +/- 0.01, N = 36.825.033.773.593.553.553.513.453.453.433.423.413.353.353.353.343.343.34MIN: 3.16 / MAX: 64.72MIN: 3.07 / MAX: 228.55MIN: 3.02 / MAX: 511.95MIN: 3.3 / MAX: 25.28MIN: 3.39 / MAX: 5.48MIN: 3.27 / MAX: 22.86MIN: 3.38 / MAX: 5.4MIN: 3.32 / MAX: 4.91MIN: 3.32 / MAX: 3.85MIN: 3.3 / MAX: 4.15MIN: 3.28 / MAX: 4.19MIN: 3.28 / MAX: 4.87MIN: 3.31 / MAX: 3.68MIN: 3.3 / MAX: 3.82MIN: 3.31 / MAX: 3.8MIN: 3.3 / MAX: 4.19MIN: 3.31 / MAX: 3.77MIN: 3.3 / MAX: 3.851. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: regnety_400m3070inv 4090RTX 3070 Ti40904080 zzz4080 xxx4080 rep4090 rep40803090gfbdae3090 repc48121620SE +/- 0.21, N = 15SE +/- 0.06, N = 3SE +/- 0.11, N = 3SE +/- 0.02, N = 316.229.889.819.078.648.588.568.568.488.398.388.368.348.218.178.168.108.028.00MIN: 7.74 / MAX: 314.84MIN: 8.14 / MAX: 251.77MIN: 7.82 / MAX: 241.19MIN: 7.61 / MAX: 402.49MIN: 8.28 / MAX: 10.42MIN: 8.13 / MAX: 9.78MIN: 8.15 / MAX: 9.8MIN: 8.17 / MAX: 10.28MIN: 8.09 / MAX: 9.64MIN: 8 / MAX: 10.29MIN: 8.31 / MAX: 8.86MIN: 8.27 / MAX: 9.08MIN: 7.99 / MAX: 26.72MIN: 8.14 / MAX: 8.84MIN: 7.99 / MAX: 8.97MIN: 7.9 / MAX: 8.99MIN: 7.98 / MAX: 8.84MIN: 7.95 / MAX: 8.63MIN: 7.94 / MAX: 8.881. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: squeezenet_ssd307040904090 rep4080 repRTX 3070 Tinv 40904080 xxx4080 zzz30903090 rep48121620SE +/- 0.23, N = 314.279.519.387.677.577.487.277.257.127.09MIN: 7.01 / MAX: 51.13MIN: 7.11 / MAX: 307.17MIN: 6.77 / MAX: 224.11MIN: 7.06 / MAX: 9.96MIN: 6.69 / MAX: 10MIN: 6.85 / MAX: 9.67MIN: 6.73 / MAX: 8.77MIN: 6.72 / MAX: 8.05MIN: 7.04 / MAX: 7.97MIN: 7.02 / MAX: 7.861. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: blazeface30704090 repnv 4090RTX 3070 Ti3090 rep40900.60531.21061.81592.42123.0265SE +/- 0.04, N = 32.691.421.401.401.381.35MIN: 1.35 / MAX: 48.81MIN: 1.36 / MAX: 2.03MIN: 1.34 / MAX: 1.86MIN: 1.28 / MAX: 1.91MIN: 1.36 / MAX: 1.73MIN: 1.28 / MAX: 1.841. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: mobilenet-v330704090RTX 3070 Ti4090 rep4080 rep4080 zzz40803090 repnv 409030904080 xxx1.34782.69564.04345.39126.739SE +/- 0.13, N = 135.993.623.443.343.313.263.263.183.173.133.05MIN: 3.05 / MAX: 26.81MIN: 3.47 / MAX: 4.24MIN: 2.65 / MAX: 361.91MIN: 3.19 / MAX: 3.99MIN: 3.16 / MAX: 3.93MIN: 3.12 / MAX: 4.74MIN: 3.13 / MAX: 4.7MIN: 3.13 / MAX: 3.61MIN: 3.04 / MAX: 4.3MIN: 3.09 / MAX: 3.68MIN: 2.94 / MAX: 3.561. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3 - Model: blazeface3070RTX 3070 Ti4080 repnv 40904090 rep30903090 rep4080 xxx4080 zzz40900.56931.13861.70792.27722.8465SE +/- 0.48, N = 32.532.481.431.421.421.391.381.321.311.30MIN: 1.08 / MAX: 118.73MIN: 1.17 / MAX: 344.52MIN: 1.36 / MAX: 2.02MIN: 1.34 / MAX: 1.99MIN: 1.34 / MAX: 2.37MIN: 1.37 / MAX: 1.48MIN: 1.35 / MAX: 1.64MIN: 1.26 / MAX: 2.03MIN: 1.25 / MAX: 1.76MIN: 1.24 / MAX: 1.921. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: mobilenet-v33070nv 4090RTX 3070 Ti4090 rep409030901.34332.68664.02995.37326.7165SE +/- 0.17, N = 155.974.813.523.443.363.16MIN: 2.84 / MAX: 111.8MIN: 3.13 / MAX: 149.75MIN: 2.95 / MAX: 536.1MIN: 3.3 / MAX: 4.34MIN: 3.21 / MAX: 4.83MIN: 3.12 / MAX: 3.671. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v2-v2 - Model: mobilenet-v230704090RTX 3070 Ti4090 repnv 40903090 rep1.3322.6643.9965.3286.66SE +/- 0.53, N = 35.924.753.833.383.293.15MIN: 3.16 / MAX: 103.24MIN: 2.93 / MAX: 147.66MIN: 3.11 / MAX: 343.21MIN: 3.2 / MAX: 4MIN: 3.12 / MAX: 4.27MIN: 3.1 / MAX: 3.751. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: shufflenet-v230704090 repRTX 3070 Ti4090nv 40903090 rep3090246810SE +/- 0.20, N = 156.305.183.923.473.373.363.36MIN: 3.28 / MAX: 147.57MIN: 3.45 / MAX: 200.36MIN: 3.12 / MAX: 496.78MIN: 3.33 / MAX: 5.01MIN: 3.25 / MAX: 5.26MIN: 3.33 / MAX: 3.83MIN: 3.32 / MAX: 3.821. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU - Model: squeezenet_ssd3070nv 4090iRTX 3070 Ti4090 rep40904080 rep40804080 xxx4080 zzz3090fg3090 repcdab3691215SE +/- 0.24, N = 15SE +/- 0.01, N = 3SE +/- 0.01, N = 313.209.118.968.137.987.867.867.737.627.557.527.237.137.127.107.097.077.06MIN: 6.9 / MAX: 68.61MIN: 6.77 / MAX: 101.58MIN: 6.92 / MAX: 244.02MIN: 6.37 / MAX: 399.11MIN: 7.32 / MAX: 16.07MIN: 7.25 / MAX: 8.98MIN: 7.22 / MAX: 10.84MIN: 7.13 / MAX: 9.7MIN: 7.01 / MAX: 8.84MIN: 7 / MAX: 8.72MIN: 7.45 / MAX: 7.74MIN: 7.15 / MAX: 8.02MIN: 7.04 / MAX: 8.43MIN: 7.05 / MAX: 7.63MIN: 7.05 / MAX: 7.65MIN: 6.99 / MAX: 9.39MIN: 7.01 / MAX: 8.07MIN: 7.01 / MAX: 7.551. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: FastestDet3070nv 409040904090 repRTX 3070 Ti3090 rep246810SE +/- 0.15, N = 37.235.864.624.594.144.07MIN: 3.75 / MAX: 121.71MIN: 3.9 / MAX: 190.17MIN: 4.48 / MAX: 5.16MIN: 4.44 / MAX: 5.2MIN: 3.73 / MAX: 5.07MIN: 4.04 / MAX: 4.251. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: shufflenet-v230704090 rep4090RTX 3070 Tinv 40903090 rep1.32532.65063.97595.30126.6265SE +/- 0.02, N = 35.895.233.563.483.433.32MIN: 3.19 / MAX: 97.88MIN: 3.34 / MAX: 185.57MIN: 3.43 / MAX: 4.24MIN: 3.33 / MAX: 5.22MIN: 3.29 / MAX: 5.31MIN: 3.29 / MAX: 3.621. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v2-v2 - Model: mobilenet-v23070RTX 3070 Ti4090nv 40904090 rep40804080 zzz4080 rep3090 rep4080 xxx30901.22852.4573.68554.9146.1425SE +/- 0.18, N = 155.463.663.483.423.363.293.283.273.173.143.12MIN: 3.27 / MAX: 38.65MIN: 2.73 / MAX: 398.42MIN: 3.32 / MAX: 4.99MIN: 3.15 / MAX: 25.1MIN: 3.17 / MAX: 4.8MIN: 3.12 / MAX: 3.99MIN: 3.11 / MAX: 4.26MIN: 3.08 / MAX: 4.68MIN: 3.12 / MAX: 3.89MIN: 3 / MAX: 3.85MIN: 3.07 / MAX: 3.621. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v2-v2 - Model: mobilenet-v230704090RTX 3070 Ti4090 repnv 40903090 rep30901.23532.47063.70594.94126.1765SE +/- 0.20, N = 155.495.253.693.603.273.173.15MIN: 2.97 / MAX: 152.08MIN: 3.11 / MAX: 367.53MIN: 3.07 / MAX: 544.13MIN: 3.44 / MAX: 4.27MIN: 3.11 / MAX: 4.1MIN: 3.12 / MAX: 3.78MIN: 3.11 / MAX: 3.781. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: efficientnet-b03070nv 40904090RTX 3070 Ti4090 rep3090 rep246810SE +/- 0.08, N = 36.635.824.634.174.103.84MIN: 3.75 / MAX: 22.34MIN: 3.98 / MAX: 197.79MIN: 4.38 / MAX: 6.01MIN: 3.86 / MAX: 5.52MIN: 3.88 / MAX: 5.04MIN: 3.8 / MAX: 4.671. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU - Model: mnasnet3070nv 4090RTX 3070 Tii40904090 rep4080 rep4080 xxx40804080 zzz3090ga3090 repfdecb1.14532.29063.43594.58125.7265SE +/- 0.16, N = 15SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 35.094.703.403.203.183.153.083.073.073.062.992.982.982.972.972.972.962.962.95MIN: 2.86 / MAX: 53.75MIN: 3 / MAX: 188.08MIN: 2.72 / MAX: 432.18MIN: 3.07 / MAX: 3.86MIN: 3.05 / MAX: 4.64MIN: 3 / MAX: 4.54MIN: 2.94 / MAX: 3.67MIN: 2.95 / MAX: 4.19MIN: 2.93 / MAX: 4.63MIN: 2.92 / MAX: 3.73MIN: 2.96 / MAX: 3.14MIN: 2.94 / MAX: 3.65MIN: 2.92 / MAX: 4.03MIN: 2.94 / MAX: 3.28MIN: 2.93 / MAX: 3.66MIN: 2.92 / MAX: 3.34MIN: 2.91 / MAX: 5.9MIN: 2.93 / MAX: 3.41MIN: 2.92 / MAX: 3.421. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3 - Model: mobilenet-v33070RTX 3070 Ti4090 repnv 40904080 xxxi40804080 zzz4090gcb3090 rep3090f1.21052.4213.63154.8426.0525SE +/- 0.22, N = 145.383.763.533.353.313.293.283.273.253.163.163.163.153.153.15MIN: 2.74 / MAX: 121.29MIN: 2.89 / MAX: 366.04MIN: 3.2 / MAX: 40.81MIN: 3.21 / MAX: 5.23MIN: 3.16 / MAX: 5.3MIN: 3.15 / MAX: 4.32MIN: 3.14 / MAX: 3.89MIN: 3.14 / MAX: 4.63MIN: 3.11 / MAX: 4.74MIN: 3.11 / MAX: 3.93MIN: 3.12 / MAX: 3.7MIN: 3.12 / MAX: 3.69MIN: 3.11 / MAX: 3.6MIN: 3.11 / MAX: 3.71MIN: 3.11 / MAX: 3.481. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: shufflenet-v23070RTX 3070 Ti4090nv 40904090 rep40804080 zzz4080 rep3090 rep4080 xxx30901.25782.51563.77345.03126.289SE +/- 0.20, N = 155.593.893.523.503.483.483.443.443.343.343.32MIN: 3.32 / MAX: 42.33MIN: 3.08 / MAX: 345.39MIN: 3.38 / MAX: 4.23MIN: 3.37 / MAX: 4.2MIN: 3.34 / MAX: 4.1MIN: 3.34 / MAX: 4.88MIN: 3.31 / MAX: 4.85MIN: 3.31 / MAX: 4.32MIN: 3.3 / MAX: 3.79MIN: 3.22 / MAX: 3.97MIN: 3.29 / MAX: 3.791. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - Model: FastestDet3070RTX 3070 Ti3090 rep3090nv 40904090 rep40901.0082.0163.0244.0325.04SE +/- 0.29, N = 154.484.254.104.084.013.912.82MIN: 2.2 / MAX: 27.6MIN: 2.46 / MAX: 526.3MIN: 4.06 / MAX: 4.2MIN: 4.04 / MAX: 4.2MIN: 3.87 / MAX: 5.47MIN: 3.77 / MAX: 5.87MIN: 2.69 / MAX: 3.51. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: mnasnet3070RTX 3070 Ti4090nv 40904090 rep4080 zzzi40804080 xxx4080 rep3090 repg30901.03282.06563.09844.13125.164SE +/- 0.14, N = 154.593.343.173.163.133.083.073.063.053.032.992.972.96MIN: 2.88 / MAX: 20.12MIN: 2.68 / MAX: 393.6MIN: 3.03 / MAX: 3.66MIN: 3.02 / MAX: 4.6MIN: 3.01 / MAX: 3.62MIN: 2.93 / MAX: 4.42MIN: 2.93 / MAX: 3.84MIN: 2.94 / MAX: 3.67MIN: 2.91 / MAX: 3.67MIN: 2.91 / MAX: 4.45MIN: 2.96 / MAX: 3.32MIN: 2.93 / MAX: 3.88MIN: 2.93 / MAX: 3.311. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3 - Model: shufflenet-v23070RTX 3070 Ti4080 xxx4080 rep40804090 rep4080 zzz3090 repig40903090fcbnv 40901.10032.20063.30094.40125.5015SE +/- 0.22, N = 154.893.953.513.443.443.403.373.363.363.353.343.343.333.333.333.17MIN: 3.04 / MAX: 18.32MIN: 3.19 / MAX: 410.41MIN: 3.37 / MAX: 4.26MIN: 3.32 / MAX: 4.16MIN: 3.31 / MAX: 4.88MIN: 3.26 / MAX: 4.84MIN: 3.25 / MAX: 3.95MIN: 3.32 / MAX: 3.7MIN: 3.25 / MAX: 4.02MIN: 3.31 / MAX: 4.01MIN: 3.23 / MAX: 4.78MIN: 3.31 / MAX: 3.6MIN: 3.29 / MAX: 3.99MIN: 3.31 / MAX: 3.81MIN: 3.3 / MAX: 3.77MIN: 3.04 / MAX: 3.781. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20230517Target: Vulkan GPU-v3-v3-v3-v3-v3-v3 - Model: blazeface3070RTX 3070 Ti4090 rep409040804080 zzz4080 xxx4080 repi3090 repg3090nv 40900.39830.79661.19491.59321.9915SE +/- 0.12, N = 141.771.491.461.431.431.421.421.411.411.391.381.361.18MIN: 1.08 / MAX: 12.53MIN: 1.05 / MAX: 379.08MIN: 1.39 / MAX: 2.91MIN: 1.36 / MAX: 2.04MIN: 1.36 / MAX: 2.06MIN: 1.34 / MAX: 2.84MIN: 1.35 / MAX: 2MIN: 1.34 / MAX: 2.1MIN: 1.35 / MAX: 2.02MIN: 1.37 / MAX: 1.52MIN: 1.35 / MAX: 2.09MIN: 1.34 / MAX: 1.46MIN: 1.11 / MAX: 1.851. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

173 Results Shown

vkpeak:
  int32-scalar
  int32-vec4
VkFFT:
  FFT + iFFT C2C 1D batched in double precision
  FFT + iFFT C2C Bluestein benchmark in double precision
vkpeak:
  fp64-vec4
  fp64-scalar
  int16-vec4
VkResample
VkFFT:
  FFT + iFFT C2C 1D batched in single precision
  FFT + iFFT C2C 1D batched in single precision, no reshuffling
  FFT + iFFT C2C 1D batched in half precision
NCNN
VkFFT:
  FFT + iFFT R2C / C2R
  FFT + iFFT C2C multidimensional in single precision
NCNN
vkpeak:
  fp32-scalar
  fp32-vec4
NCNN
vkpeak
NCNN
vkpeak:
  fp16-vec4
  int16-scalar
VkResample
NCNN:
  Vulkan GPU-v3-v3-v3 - blazeface
  CPU-v3-v3-v3-v2-v2 - mobilenet-v2
  Vulkan GPU - blazeface
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mnasnet
  CPU - mobilenet
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - alexnet
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mnasnet
VkFFT
NCNN:
  CPU - resnet18
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - squeezenet_ssd
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - alexnet
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v2-v2 - mobilenet-v2
  Vulkan GPU-v3-v3-v3 - googlenet
  Vulkan GPU - mobilenet
  CPU-v3-v3-v3 - blazeface
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - blazeface
  CPU-v3-v3-v3 - vision_transformer
  CPU-v3-v3-v3 - alexnet
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - resnet18
  Vulkan GPU-v3-v3-v3 - resnet18
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - efficientnet-b0
  CPU-v3-v3-v3 - squeezenet_ssd
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - FastestDet
  Vulkan GPU - alexnet
  CPU - mnasnet
  Vulkan GPU - googlenet
  CPU-v3-v3-v3-v3-v3-v3 - FastestDet
  Vulkan GPU-v2-v2 - mobilenet-v2
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - alexnet
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - efficientnet-b0
  CPU-v3-v3-v3 - regnety_400m
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - FastestDet
  CPU-v3-v3-v3-v3-v3-v3 - alexnet
  CPU-v3-v3-v3 - shufflenet-v2
  Vulkan GPU - resnet18
  CPU-v3-v3-v3-v3-v3-v3 - resnet18
  CPU-v3-v3-v3-v3-v3-v3 - googlenet
  CPU - vgg16
  CPU-v3-v3-v3 - resnet50
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - shufflenet-v2
  CPU - efficientnet-b0
  CPU-v3-v3 - mobilenet-v3
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - googlenet
  CPU-v3-v3-v3-v3-v3-v3 - efficientnet-b0
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - googlenet
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - blazeface
  CPU - vision_transformer
  Vulkan GPU-v3-v3-v3 - mnasnet
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - googlenet
  Vulkan GPU-v3-v3-v3 - vgg16
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - vgg16
  CPU-v3-v3-v3 - efficientnet-b0
  Vulkan GPU - efficientnet-b0
  CPU-v3-v3-v3-v3-v3-v3 - shufflenet-v2
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - resnet50
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - resnet50
  Vulkan GPU - resnet50
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - resnet50
  CPU - blazeface
  Vulkan GPU-v3-v3-v3 - resnet50
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - resnet18
  Vulkan GPU-v3-v3-v3 - FastestDet
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - alexnet
  CPU - googlenet
  CPU-v3-v3-v3 - mnasnet
  Vulkan GPU-v3-v3-v3 - vision_transformer
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - yolov4-tiny
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenet
  Vulkan GPU-v3-v3-v3 - yolov4-tiny
  Vulkan GPU-v3-v3-v3-v2-v2 - mobilenet-v2
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - resnet18
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenet-v3
  Vulkan GPU-v3-v3-v3 - squeezenet_ssd
  Vulkan GPU-v3-v3-v3 - alexnet
  CPU-v3-v3-v3 - yolov4-tiny
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - mobilenet
  CPU-v3-v3-v3-v3-v3-v3-v2-v2 - mobilenet-v2
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - yolov4-tiny
  CPU-v3-v3-v3-v3-v3-v3 - regnety_400m
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - vgg16
  Vulkan GPU - squeezenet_ssd
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - regnety_400m
  CPU - alexnet
  Vulkan GPU - yolov4-tiny
  Vulkan GPU-v3-v3-v3 - regnety_400m
  CPU - regnety_400m
  CPU-v3-v3-v3 - mobilenet
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - yolov4-tiny
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - regnety_400m
  Vulkan GPU - vision_transformer
  CPU-v3-v3-v3-v3-v3-v3 - yolov4-tiny
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - vision_transformer
  CPU-v3-v3-v3-v3-v3-v3 - resnet50
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - vision_transformer
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - vision_transformer
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - efficientnet-b0
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - squeezenet_ssd
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - squeezenet_ssd
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - vgg16
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - regnety_400m
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - resnet18
  CPU-v3-v3-v3 - googlenet
  CPU - yolov4-tiny
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - squeezenet_ssd
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - googlenet
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - vision_transformer
  CPU - resnet50
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - resnet50
  CPU-v3-v3-v3-v3-v3-v3 - vgg16
  CPU-v3-v3-v3 - resnet18
  Vulkan GPU - FastestDet
  Vulkan GPU - shufflenet-v2
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenet
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - vgg16
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenet
  CPU-v3-v3-v3 - vgg16
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - regnety_400m
  Vulkan GPU-v3-v3 - mobilenet-v3
  CPU-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenet-v3
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenet-v3
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mnasnet
  Vulkan GPU - vgg16
  Vulkan GPU-v3-v3-v3 - mobilenet
  CPU-v3-v3-v3-v3-v3-v3 - vision_transformer
  CPU-v3-v3-v3-v3-v3-v3 - mnasnet
  CPU-v3-v3-v3-v3-v3-v3 - mobilenet
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - yolov4-tiny
  Vulkan GPU-v3-v3-v3 - efficientnet-b0
  CPU - shufflenet-v2
  Vulkan GPU - regnety_400m
  CPU-v3-v3-v3-v3-v3-v3 - squeezenet_ssd
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - blazeface
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenet-v3
  CPU-v3-v3-v3-v3-v3-v3 - blazeface
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - mobilenet-v3
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v2-v2 - mobilenet-v2
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - shufflenet-v2
  CPU - squeezenet_ssd
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - FastestDet
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - shufflenet-v2
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v2-v2 - mobilenet-v2
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v2-v2 - mobilenet-v2
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3-v3 - efficientnet-b0
  Vulkan GPU - mnasnet
  CPU-v3-v3-v3-v3-v3 - mobilenet-v3
  Vulkan GPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - shufflenet-v2
  CPU-v3-v3-v3-v3-v3-v3-v3-v3-v3 - FastestDet
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - mnasnet
  Vulkan GPU-v3-v3-v3 - shufflenet-v2
  Vulkan GPU-v3-v3-v3-v3-v3-v3 - blazeface