neo icon indicating copy to clipboard operation
neo copied to clipboard

Published 20 hours ago • verified publisher icon andreaferretti

Implement row-major CUDA matrices

Open andreaferretti opened this issue 8 years ago • 1 comments

The relevant operations in CUBLAS (gesv, gemm, geam) feature a CUBLAS_T flag to work with transposed matrices. This should be enough to cover matrices in row order

andreaferretti avatar Jun 13 '17 16:06 andreaferretti

This would also allow to support t() in the CUDA version

andreaferretti avatar Jun 13 '17 16:06 andreaferretti