Open discussion for MAGMA library (Matrix Algebra on GPU and Multicore Architectures)
-
cudacoder
- Posts: 2
- Joined: Thu May 12, 2011 5:24 am
Post
by cudacoder » Thu May 12, 2011 5:33 am
Hello,
is it possible to call MAGMA library functions directly from __global__ or __device__ kernels in CUDA?
Best regards