I am new in CUDA and trying to understand the boxfilter example in SDK. Can anyone help me with the following line of codes:
d_boxfilter_rgba_x<<< height / nthreads, nthreads, 0 >>>( d_temp, width, height, radius);
what is <<< height / nthreads, nthreads, 0 >>> doing there? Is it a template setting up the block and thread?