ncnn test on x86

[yuxiaoyuan0406]

Platform Information

Item	Value
Platform	GNU/Linux 20.04.1-Ubuntu
CPU Model	Intel(R) Core(TM) i7-10710U CPU @ 1.10GHz
Architecture	x86_64
$(nproc)	12

Benchmark Result

ref:
Benchmark on linux
Benchmark

$ ../build/benchmark/benchncnn 10 $(nproc) 0 0
[0 Intel(R) UHD Graphics (CML GT2)]  queueC=0[1]  queueG=0[1]  queueT=0[1]
[0 Intel(R) UHD Graphics (CML GT2)]  bugsbn1=0  bugbilz=0  bugcopc=0  bugihfa=0
[0 Intel(R) UHD Graphics (CML GT2)]  fp16-p/s/a=1/1/1  int8-p/s/a=1/1/1
[0 Intel(R) UHD Graphics (CML GT2)]  subgroup=32  basic=1  vote=1  ballot=1  shuffle=1
[1 llvmpipe (LLVM 12.0.0, 256 bits)]  queueC=0[1]  queueG=0[1]  queueT=0[1]
[1 llvmpipe (LLVM 12.0.0, 256 bits)]  bugsbn1=0  bugbilz=0  bugcopc=0  bugihfa=0
[1 llvmpipe (LLVM 12.0.0, 256 bits)]  fp16-p/s/a=1/1/0  int8-p/s/a=1/1/0
[1 llvmpipe (LLVM 12.0.0, 256 bits)]  subgroup=8  basic=1  vote=1  ballot=1  shuffle=0
loop_count = 10
num_threads = 12
powersave = 0
gpu_device = 0
cooling_down = 1
          squeezenet  min =   16.91  max =   17.67  avg =   17.32
     squeezenet_int8  min =    7.87  max =   11.48  avg =    8.51
           mobilenet  min =    9.36  max =   18.25  avg =   11.42
      mobilenet_int8  min =    7.87  max =   13.09  avg =    9.33
        mobilenet_v2  min =   17.55  max =   18.56  avg =   17.96
        mobilenet_v3  min =   18.43  max =   19.73  avg =   19.11
          shufflenet  min =   17.74  max =   19.79  avg =   18.03
       shufflenet_v2  min =   15.85  max =   17.99  avg =   16.47
             mnasnet  min =   11.33  max =   13.98  avg =   12.84
     proxylessnasnet  min =   17.63  max =   20.01  avg =   18.05
     efficientnet_b0  min =   20.87  max =   23.90  avg =   22.76
   efficientnetv2_b0  min =   34.38  max =   36.47  avg =   34.97
        regnety_400m  min =   18.35  max =   21.97  avg =   19.27
           blazeface  min =    4.86  max =    9.09  avg =    5.54
           googlenet  min =   21.29  max =   22.96  avg =   21.55
      googlenet_int8  min =   25.77  max =   36.60  avg =   28.39
            resnet18  min =   22.48  max =   23.21  avg =   22.81
       resnet18_int8  min =   26.86  max =   27.15  avg =   26.98
             alexnet  min =   20.38  max =   21.29  avg =   20.60
               vgg16  min =  120.37  max =  202.21  avg =  164.36
          vgg16_int8  min =   86.16  max =   89.87  avg =   87.00
            resnet50  min =   46.53  max =   47.06  avg =   46.82
       resnet50_int8  min =   49.62  max =   53.48  avg =   50.32
      squeezenet_ssd  min =   39.04  max =   56.94  avg =   45.01
 squeezenet_ssd_int8  min =   20.88  max =   26.68  avg =   22.49
       mobilenet_ssd  min =   18.43  max =   19.24  avg =   18.81
  mobilenet_ssd_int8  min =   16.86  max =   30.28  avg =   21.73
      mobilenet_yolo  min =   36.53  max =   40.75  avg =   37.95
  mobilenetv2_yolov3  min =   23.98  max =   43.53  avg =   29.55
         yolov4-tiny  min =   52.50  max =   89.87  avg =   72.60
           nanodet_m  min =   14.63  max =   19.39  avg =   16.37
    yolo-fastest-1.1  min =    7.19  max =   16.75  avg =    8.25
      yolo-fastestv2  min =   17.56  max =   18.14  avg =   17.90

[yuxiaoyuan0406]

Platform Information

Run another test on jetson nano

Item	Value
Platform	GNU/Linux Ubuntu
CPU	Cortex-A57
Architecture	aarch64
$(nproc)	4

Benchmark result

command

../build/benchmark/benchncnn 10 $(nproc) 0 0 / 2> ../build/benchmark/result_release.txt

[0 NVIDIA Tegra X1 (nvgpu)]  queueC=0[16]  queueG=0[16]  queueT=0[16]
[0 NVIDIA Tegra X1 (nvgpu)]  bugsbn1=0  bugbilz=0  bugcopc=0  bugihfa=0
[0 NVIDIA Tegra X1 (nvgpu)]  fp16-p/s/a=1/1/1  int8-p/s/a=1/1/1
[0 NVIDIA Tegra X1 (nvgpu)]  subgroup=32  basic=1  vote=1  ballot=1  shuffle=1
loop_count = 10
num_threads = 4
powersave = 0
gpu_device = 0
cooling_down = 0
          squeezenet  min =   28.29  max =   31.14  avg =   29.85
     squeezenet_int8  min =   30.52  max =   31.51  avg =   30.77
           mobilenet  min =   24.36  max =   40.65  avg =   32.87
      mobilenet_int8  min =   32.00  max =   40.29  avg =   33.22
        mobilenet_v2  min =   23.52  max =   28.88  avg =   26.46
        mobilenet_v3  min =   34.13  max =   40.69  avg =   38.28
          shufflenet  min =   22.02  max =   25.63  avg =   23.11
       shufflenet_v2  min =   25.83  max =   31.35  avg =   27.61
             mnasnet  min =   33.61  max =   41.15  avg =   35.16
     proxylessnasnet  min =   18.20  max =   21.11  avg =   19.46
     efficientnet_b0  min =   29.73  max =   38.72  avg =   32.85
   efficientnetv2_b0  min =   68.96  max =  104.47  avg =   90.10
        regnety_400m  min =   32.92  max =   36.11  avg =   34.39
           blazeface  min =   20.55  max =   22.88  avg =   21.38
           googlenet  min =   37.13  max =   59.34  avg =   49.02
      googlenet_int8  min =   90.09  max =   92.39  avg =   90.90
            resnet18  min =   29.43  max =   38.27  avg =   32.38
       resnet18_int8  min =   65.44  max =   73.70  avg =   66.81
             alexnet  min =   49.54  max =   52.00  avg =   50.51
               vgg16  min =  146.21  max =  150.37  avg =  148.23
          vgg16_int8  min =  417.62  max =  571.94  avg =  487.39
            resnet50  min =   74.66  max =   88.65  avg =   77.93
       resnet50_int8  min =  186.79  max =  189.20  avg =  188.06
      squeezenet_ssd  min =   38.81  max =   51.99  avg =   43.81
 squeezenet_ssd_int8  min =   82.06  max =  112.05  avg =   89.67
       mobilenet_ssd  min =   44.51  max =   48.70  avg =   46.88
  mobilenet_ssd_int8  min =   69.47  max =  319.72  avg =  109.66
      mobilenet_yolo  min =   74.73  max =   84.99  avg =   80.13
  mobilenetv2_yolov3  min =   52.57  max =   58.36  avg =   55.76
         yolov4-tiny  min =   71.71  max =   77.73  avg =   74.80
           nanodet_m  min =   37.31  max =   47.64  avg =   40.79
    yolo-fastest-1.1  min =   29.69  max =   41.20  avg =   35.88
      yolo-fastestv2  min =   25.56  max =   30.13  avg =   27.89

full result

[smy1999]

Platform Information

Item	Value
Platform	macOS Monterey 12.3.1
CPU Model	Apple M1 Pro
Architecture	ARMv8-A
$(nproc)	16

Benchmark Result

ref:
Benchmark on macOS
Benchmark

smy1999@smy-mbp benchmark % ./benchncnn 10 1 0 -1 0
loop_count = 10
num_threads = 1
powersave = 0
gpu_device = -1
cooling_down = 0
          squeezenet  min =    4.77  max =    5.20  avg =    4.91
     squeezenet_int8  min =    4.00  max =    4.30  avg =    4.08
           mobilenet  min =    8.70  max =    8.77  avg =    8.74
      mobilenet_int8  min =    4.61  max =    4.87  avg =    4.69
        mobilenet_v2  min =    5.35  max =    5.53  avg =    5.38
        mobilenet_v3  min =    4.42  max =    4.48  avg =    4.44
          shufflenet  min =    3.35  max =    3.82  avg =    3.46
       shufflenet_v2  min =    3.66  max =    3.72  avg =    3.67
             mnasnet  min =    5.42  max =    5.47  avg =    5.44
     proxylessnasnet  min =    6.91  max =    6.97  avg =    6.93
     efficientnet_b0  min =    8.67  max =    8.83  avg =    8.71
   efficientnetv2_b0  min =   16.79  max =   17.33  avg =   16.95
        regnety_400m  min =    7.37  max =    8.17  avg =    7.67
           blazeface  min =    1.82  max =    1.99  avg =    1.85
           googlenet  min =   25.02  max =   25.40  avg =   25.13
      googlenet_int8  min =   16.60  max =   17.12  avg =   16.71
            resnet18  min =   16.90  max =   17.32  avg =   17.02
       resnet18_int8  min =   15.12  max =   15.24  avg =   15.15
             alexnet  min =   30.06  max =   30.89  avg =   30.22
               vgg16  min =   71.93  max =   73.46  avg =   72.75
          vgg16_int8  min =   99.24  max =  100.04  avg =   99.62
            resnet50  min =   43.33  max =   45.81  avg =   43.73
       resnet50_int8  min =   29.60  max =   29.88  avg =   29.69
      squeezenet_ssd  min =   20.89  max =   21.10  avg =   21.00
 squeezenet_ssd_int8  min =   18.92  max =   19.26  avg =   19.03
       mobilenet_ssd  min =   21.86  max =   29.37  avg =   22.76
  mobilenet_ssd_int8  min =   13.09  max =   13.38  avg =   13.18
      mobilenet_yolo  min =   40.09  max =   40.44  avg =   40.26
  mobilenetv2_yolov3  min =   20.66  max =   27.30  avg =   21.80
         yolov4-tiny  min =   30.76  max =   32.04  avg =   31.44
           nanodet_m  min =    9.10  max =    9.93  avg =    9.43
    yolo-fastest-1.1  min =    3.34  max =    3.57  avg =    3.41
      yolo-fastestv2  min =    3.39  max =    3.52  avg =    3.42