Hey all together!
I have a little question about the ‘grid’. What is the advantage to use different grid dimension for my kernel function? I thought there can be only one kernel be aktive on the gpu everytime, so it can use all the grid, can’t it? So, what is the sense of defining a special grid dimension?
Sorry, because of this stupid question (and my bad english). I’m still new in CUDA.
Thanks in advance!