Skip to content

Manage cuda device memory with cudaMalloc instead of thrust::vector.

Closes #114 (closed)

Merge request reports