Hello. I have tested perfomance of cpu and gpu on tx2 using these simple tests on opencv:
for gpu:
auto canny_edg = cv::cuda::createCannyEdgeDetector(50, 150, 3);
for (int j = 0; j < 100; j++)
{
timer.start();
cv::cuda::GpuMat canny_output;
canny_edg->detect(frame, canny_output);
qDebug() << timer.elapsed();
}
some part of time sample (ms):
13
13
12
14
14
12
12
13
and for cpu:
for (int j = 0; j < 100; j++)
{
timer.start();
cv::Mat canny_output;
cv::Canny(flipped, canny_output, 50, 150, 3);
qDebug() << timer.elapsed();
}
some part of time sample (ms):
8
10
11
12
19
7
11
12
as we can see, results are same in average. is it ok? i have expected some better perfomance on gpu. probably i have missed some gpu options?
my sudo ./jetson_clocks.sh --show:
SOC family:tegra186 Machine:quill
Online CPUs: 0,3-5
CPU Cluster Switching: Disabled
cpu0: Gonvernor=schedutil MinFreq=2035200 MaxFreq=2035200 CurrentFreq=2035200
cpu1: Gonvernor=schedutil MinFreq=345600 MaxFreq=2035200 CurrentFreq=2035200
cpu2: Gonvernor=schedutil MinFreq=345600 MaxFreq=2035200 CurrentFreq=2035200
cpu3: Gonvernor=schedutil MinFreq=2035200 MaxFreq=2035200 CurrentFreq=2035200
cpu4: Gonvernor=schedutil MinFreq=2035200 MaxFreq=2035200 CurrentFreq=2035200
cpu5: Gonvernor=schedutil MinFreq=2035200 MaxFreq=2035200 CurrentFreq=2035200
GPU MinFreq=1122000000 MaxFreq=1122000000 CurrentFreq=1122000000
EMC MinFreq=40800000 MaxFreq=1600000000 CurrentFreq=1600000000 FreqOverride=1
Fan: speed=255