News

Grayscale optimization

Added by Remaldeep Singh about 1 year ago

Well it is a very strange bug but cudaMemcpy for 320*240*4 works faster than 320*240*1. Maybe because of half warp allignment????

Also available in: Atom