matlab - How to multiply many gpuArray matrices in parallel? -
given matrices a1,...,an , b1,...,bn stored gpuarray
, want calculate matrices ci=ai*bi.
all ai's of same size, , bi's of same (possibly different) size.
how do fast on gpu, assuming n large , sizes of matrices relatively small? possible avoid using cuda?
if have matlab r2013b, can use new gpuarray pagefun
function.
Comments
Post a Comment