ncnn mnn 5900HX

AMD Ryzen 9 5900HX testing with a ASUS G513QY v1.0 (G513QY.318 BIOS) and ASUS AMD Cezanne 512MB on Arch rolling via the Phoronix Test Suite.

HTML result view exported from: https://openbenchmarking.org/result/2208131-NE-NCNNMNN5973&grt.

ncnn mnn 5900HXProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerOpenGLVulkanCompilerFile-SystemScreen ResolutionABCAMD Ryzen 9 5900HX @ 3.30GHz (8 Cores / 16 Threads)ASUS G513QY v1.0 (G513QY.318 BIOS)AMD Renoir/Cezanne16GB512GB SAMSUNG MZVLQ512HBLU-00B00ASUS AMD Cezanne 512MB (2500/1000MHz)AMD Navi 21/23LQ156M1JW25Realtek RTL8111/8168/8411 + MEDIATEK MT7921 802.11ax PCIArch rolling5.18.16-arch1-1 (x86_64)KDE Plasma 5.25.4X Server 1.21.1.4 + Wayland4.6 Mesa 22.1.4 (LLVM 14.0.6 DRM 3.46)1.3.211GCC 12.1.1 20220730ext41920x1080ASUS AMD Cezanne 512MBASUS AMD Cezanne 512MB (2500/1000MHz)OpenBenchmarking.orgKernel Details- Transparent Huge Pages: alwaysCompiler Details- --disable-libssp --disable-libstdcxx-pch --disable-werror --enable-__cxa_atexit --enable-bootstrap --enable-cet=auto --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-default-ssp --enable-gnu-indirect-function --enable-gnu-unique-object --enable-languages=c,c++,ada,fortran,go,lto,objc,obj-c++,d --enable-libstdcxx-backtrace --enable-link-serialization=1 --enable-lto --enable-multilib --enable-plugin --enable-shared --enable-threads=posix --mandir=/usr/share/man --with-build-config=bootstrap-lto --with-linker-hash-style=gnu Processor Details- Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - Platform Profile: balanced - CPU Microcode: 0xa50000c - ACPI Profile: balanced Security 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected

ncnn mnn 5900HXmnn: mobilenetV3mnn: squeezenetv1.1mnn: resnet-v2-50mnn: SqueezeNetV1.0mnn: MobileNetV2_224mnn: mobilenet-v1-1.0mnn: inception-v3ncnn: CPU - mobilenetncnn: CPU-v2-v2 - mobilenet-v2ncnn: CPU-v3-v3 - mobilenet-v3ncnn: CPU - shufflenet-v2ncnn: CPU - mnasnetncnn: CPU - efficientnet-b0ncnn: CPU - blazefacencnn: CPU - googlenetncnn: CPU - vgg16ncnn: CPU - resnet18ncnn: CPU - alexnetncnn: CPU - resnet50ncnn: CPU - yolov4-tinyncnn: CPU - squeezenet_ssdncnn: CPU - regnety_400mncnn: CPU - vision_transformerncnn: CPU - FastestDetncnn: Vulkan GPU - mobilenetncnn: Vulkan GPU-v2-v2 - mobilenet-v2ncnn: Vulkan GPU-v3-v3 - mobilenet-v3ncnn: Vulkan GPU - shufflenet-v2ncnn: Vulkan GPU - mnasnetncnn: Vulkan GPU - efficientnet-b0ncnn: Vulkan GPU - blazefacencnn: Vulkan GPU - googlenetncnn: Vulkan GPU - vgg16ncnn: Vulkan GPU - resnet18ncnn: Vulkan GPU - alexnetncnn: Vulkan GPU - yolov4-tinyncnn: Vulkan GPU - squeezenet_ssdncnn: Vulkan GPU - regnety_400mncnn: Vulkan GPU - vision_transformerncnn: Vulkan GPU - FastestDetncnn: Vulkan GPU - resnet50ABC1.1362.38221.2604.4632.4902.36928.00414.984.173.072.252.815.780.7713.9673.1112.4510.6722.9527.4220.877.12165.652.574.481.882.301.771.895.021.073.797.182.331.666.674.053.00147.831.865.061.1372.41021.3554.5232.4432.35128.45215.244.193.012.212.735.780.7714.0173.0512.6710.8223.1227.4620.947.10167.992.604.481.922.341.821.935.061.113.947.212.31.656.654.083.03146.001.915.061.1412.40521.2264.5062.4712.33428.78815.304.143.022.282.725.750.7614.0072.9312.5410.6522.8027.5920.947.08165.922.694.491.922.351.831.945.091.113.967.222.321.666.683.983.06145.821.915.08OpenBenchmarking.org

Mobile Neural Network

Model: mobilenetV3

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.0Model: mobilenetV3ABC0.25670.51340.77011.02681.2835SE +/- 0.004, N = 3SE +/- 0.004, N = 3SE +/- 0.008, N = 31.1361.1371.141MIN: 1.12 / MAX: 2.75MIN: 1.12 / MAX: 2.74MIN: 1.12 / MAX: 2.741. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Mobile Neural Network

Model: squeezenetv1.1

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.0Model: squeezenetv1.1ABC0.54231.08461.62692.16922.7115SE +/- 0.027, N = 3SE +/- 0.021, N = 3SE +/- 0.001, N = 32.3822.4102.405MIN: 2.16 / MAX: 4.13MIN: 2.19 / MAX: 4.13MIN: 2.2 / MAX: 4.471. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Mobile Neural Network

Model: resnet-v2-50

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.0Model: resnet-v2-50ABC510152025SE +/- 0.10, N = 3SE +/- 0.02, N = 3SE +/- 0.13, N = 321.2621.3621.23MIN: 19.91 / MAX: 24.58MIN: 19.93 / MAX: 26.59MIN: 19.87 / MAX: 28.061. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Mobile Neural Network

Model: SqueezeNetV1.0

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.0Model: SqueezeNetV1.0ABC1.01772.03543.05314.07085.0885SE +/- 0.051, N = 3SE +/- 0.039, N = 3SE +/- 0.014, N = 34.4634.5234.506MIN: 4.13 / MAX: 6.63MIN: 4.23 / MAX: 7MIN: 4.21 / MAX: 6.461. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Mobile Neural Network

Model: MobileNetV2_224

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.0Model: MobileNetV2_224ABC0.56031.12061.68092.24122.8015SE +/- 0.018, N = 3SE +/- 0.024, N = 3SE +/- 0.008, N = 32.4902.4432.471MIN: 2.2 / MAX: 6.31MIN: 2.16 / MAX: 5.14MIN: 2.23 / MAX: 4.691. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Mobile Neural Network

Model: mobilenet-v1-1.0

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.0Model: mobilenet-v1-1.0ABC0.5331.0661.5992.1322.665SE +/- 0.004, N = 3SE +/- 0.018, N = 3SE +/- 0.035, N = 32.3692.3512.334MIN: 2.2 / MAX: 4.92MIN: 2.18 / MAX: 4.33MIN: 2.18 / MAX: 4.511. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Mobile Neural Network

Model: inception-v3

OpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 2.0Model: inception-v3ABC714212835SE +/- 0.05, N = 3SE +/- 0.15, N = 3SE +/- 0.29, N = 328.0028.4528.79MIN: 26.69 / MAX: 32.71MIN: 26.85 / MAX: 35.34MIN: 27.03 / MAX: 33.351. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

NCNN

Target: CPU - Model: mobilenet

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: mobilenetABC48121620SE +/- 0.05, N = 3SE +/- 0.22, N = 3SE +/- 0.21, N = 314.9815.2415.30MIN: 14.64 / MAX: 17.32MIN: 14.69 / MAX: 17.39MIN: 14.68 / MAX: 17.721. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU-v2-v2 - Model: mobilenet-v2

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU-v2-v2 - Model: mobilenet-v2ABC0.94281.88562.82843.77124.714SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 34.174.194.14MIN: 3.95 / MAX: 6.1MIN: 3.95 / MAX: 4.8MIN: 3.94 / MAX: 5.911. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU-v3-v3 - Model: mobilenet-v3

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU-v3-v3 - Model: mobilenet-v3ABC0.69081.38162.07242.76323.454SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.05, N = 33.073.013.02MIN: 2.9 / MAX: 4.97MIN: 2.86 / MAX: 3.58MIN: 2.84 / MAX: 4.661. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: shufflenet-v2

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: shufflenet-v2ABC0.5131.0261.5392.0522.565SE +/- 0.02, N = 3SE +/- 0.02, N = 3SE +/- 0.05, N = 32.252.212.28MIN: 2.13 / MAX: 3.58MIN: 2.14 / MAX: 2.38MIN: 2.12 / MAX: 3.581. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: mnasnet

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: mnasnetABC0.63231.26461.89692.52923.1615SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 32.812.732.72MIN: 2.61 / MAX: 4.72MIN: 2.61 / MAX: 3.88MIN: 2.57 / MAX: 4.391. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: efficientnet-b0

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: efficientnet-b0ABC1.30052.6013.90155.2026.5025SE +/- 0.03, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 35.785.785.75MIN: 5.56 / MAX: 7.73MIN: 5.57 / MAX: 6.46MIN: 5.56 / MAX: 7.721. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: blazeface

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: blazefaceABC0.17330.34660.51990.69320.8665SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 30.770.770.76MIN: 0.75 / MAX: 2.04MIN: 0.75 / MAX: 0.95MIN: 0.74 / MAX: 0.931. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: googlenet

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: googlenetABC48121620SE +/- 0.02, N = 3SE +/- 0.06, N = 3SE +/- 0.06, N = 313.9614.0114.00MIN: 13.58 / MAX: 15.87MIN: 13.59 / MAX: 15.88MIN: 13.57 / MAX: 16.581. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: vgg16

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: vgg16ABC1632486480SE +/- 0.07, N = 3SE +/- 0.10, N = 3SE +/- 0.04, N = 373.1173.0572.93MIN: 72.51 / MAX: 88.36MIN: 72.57 / MAX: 75.27MIN: 72.43 / MAX: 75.611. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: resnet18

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: resnet18ABC3691215SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.04, N = 312.4512.6712.54MIN: 12.16 / MAX: 14.15MIN: 12.43 / MAX: 14.75MIN: 12.23 / MAX: 14.031. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: alexnet

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: alexnetABC3691215SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.00, N = 310.6710.8210.65MIN: 10.41 / MAX: 11.91MIN: 10.58 / MAX: 11.12MIN: 10.36 / MAX: 11.271. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: resnet50

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: resnet50ABC612182430SE +/- 0.06, N = 3SE +/- 0.13, N = 3SE +/- 0.04, N = 322.9523.1222.80MIN: 22.25 / MAX: 27.03MIN: 22.33 / MAX: 25.11MIN: 22.18 / MAX: 25.031. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: yolov4-tiny

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: yolov4-tinyABC612182430SE +/- 0.03, N = 3SE +/- 0.12, N = 3SE +/- 0.11, N = 327.4227.4627.59MIN: 26.95 / MAX: 31.42MIN: 26.83 / MAX: 29.49MIN: 27.04 / MAX: 29.891. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: squeezenet_ssd

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: squeezenet_ssdABC510152025SE +/- 0.02, N = 3SE +/- 0.06, N = 3SE +/- 0.14, N = 320.8720.9420.94MIN: 20.53 / MAX: 22.66MIN: 20.53 / MAX: 23.04MIN: 20.47 / MAX: 22.831. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: regnety_400m

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: regnety_400mABC246810SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.03, N = 37.127.107.08MIN: 6.8 / MAX: 9.14MIN: 6.81 / MAX: 8.28MIN: 6.76 / MAX: 9.431. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: vision_transformer

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: vision_transformerABC4080120160200SE +/- 0.19, N = 3SE +/- 1.40, N = 3SE +/- 0.19, N = 3165.65167.99165.92MIN: 161.51 / MAX: 171.84MIN: 162.38 / MAX: 182.04MIN: 162.68 / MAX: 172.941. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: CPU - Model: FastestDet

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: CPU - Model: FastestDetABC0.60531.21061.81592.42123.0265SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.06, N = 32.572.602.69MIN: 2.45 / MAX: 4.5MIN: 2.48 / MAX: 2.85MIN: 2.47 / MAX: 2.931. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: mobilenet

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: mobilenetABC1.01032.02063.03094.04125.0515SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 34.484.484.49MIN: 4.41 / MAX: 6.04MIN: 4.39 / MAX: 5.72MIN: 4.42 / MAX: 5.821. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2ABC0.4320.8641.2961.7282.16SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 31.881.921.92MIN: 1.82 / MAX: 3.01MIN: 1.78 / MAX: 2.93MIN: 1.81 / MAX: 3.171. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3ABC0.52881.05761.58642.11522.644SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 32.302.342.35MIN: 2.25 / MAX: 3.64MIN: 2.23 / MAX: 2.78MIN: 2.23 / MAX: 3.681. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: shufflenet-v2

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: shufflenet-v2ABC0.41180.82361.23541.64722.059SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 31.771.821.83MIN: 1.72 / MAX: 2.71MIN: 1.73 / MAX: 3.21MIN: 1.75 / MAX: 2.831. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: mnasnet

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: mnasnetABC0.43650.8731.30951.7462.1825SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 31.891.931.94MIN: 1.83 / MAX: 3.17MIN: 1.82 / MAX: 3.16MIN: 1.8 / MAX: 3.081. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: efficientnet-b0

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: efficientnet-b0ABC1.14532.29063.43594.58125.7265SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 35.025.065.09MIN: 4.95 / MAX: 5.24MIN: 4.94 / MAX: 7.39MIN: 4.96 / MAX: 9.351. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: blazeface

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: blazefaceABC0.24980.49960.74940.99921.249SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 31.071.111.11MIN: 1 / MAX: 1.36MIN: 0.99 / MAX: 1.4MIN: 1 / MAX: 1.291. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: googlenet

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: googlenetABC0.8911.7822.6733.5644.455SE +/- 0.10, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 33.793.943.96MIN: 3.55 / MAX: 4.51MIN: 3.84 / MAX: 4.16MIN: 3.82 / MAX: 4.231. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: vgg16

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: vgg16ABC246810SE +/- 0.06, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 37.187.217.22MIN: 6.68 / MAX: 8.37MIN: 6.7 / MAX: 11.48MIN: 6.68 / MAX: 15.041. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: resnet18

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: resnet18ABC0.52431.04861.57292.09722.6215SE +/- 0.03, N = 2SE +/- 0.00, N = 3SE +/- 0.02, N = 32.332.302.32MIN: 2.24 / MAX: 3.45MIN: 2.27 / MAX: 2.56MIN: 2.27 / MAX: 3.531. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: alexnet

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: alexnetABC0.37350.7471.12051.4941.8675SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 31.661.651.66MIN: 1.6 / MAX: 2.65MIN: 1.61 / MAX: 1.94MIN: 1.61 / MAX: 2.881. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: yolov4-tiny

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: yolov4-tinyABC246810SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 36.676.656.68MIN: 6.58 / MAX: 7.31MIN: 6.58 / MAX: 7.04MIN: 6.56 / MAX: 8.071. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: squeezenet_ssd

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: squeezenet_ssdABC0.9181.8362.7543.6724.59SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.14, N = 34.054.083.98MIN: 3.93 / MAX: 4.48MIN: 3.96 / MAX: 5.28MIN: 3.65 / MAX: 5.451. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: regnety_400m

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: regnety_400mABC0.68851.3772.06552.7543.4425SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 33.003.033.06MIN: 2.93 / MAX: 3.17MIN: 2.96 / MAX: 3.25MIN: 2.93 / MAX: 3.651. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: vision_transformer

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: vision_transformerABC306090120150SE +/- 0.62, N = 3SE +/- 0.24, N = 3SE +/- 0.23, N = 3147.83146.00145.82MIN: 143.13 / MAX: 154.5MIN: 143.21 / MAX: 150.77MIN: 142.07 / MAX: 152.311. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: FastestDet

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: FastestDetABC0.42980.85961.28941.71922.149SE +/- 0.01, N = 3SE +/- 0.01, N = 2SE +/- 0.00, N = 31.861.911.91MIN: 1.81 / MAX: 3.74MIN: 1.82 / MAX: 2.06MIN: 1.82 / MAX: 2.321. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

NCNN

Target: Vulkan GPU - Model: resnet50

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20220729Target: Vulkan GPU - Model: resnet50ABC1.1432.2863.4294.5725.715SE +/- 0.00, N = 2SE +/- 0.02, N = 2SE +/- 0.00, N = 25.065.065.08MIN: 4.99 / MAX: 7.58MIN: 4.99 / MAX: 5.21MIN: 5 / MAX: 6.081. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

Phoronix Test Suite v10.8.4