Mahesh Doijade
|
067cb65523
|
-- Add partitioned cuda pipeline prod-cons gemm kernel
-- Add cudaCompressibleMemory sample to use copy engine vs SM writes
depending on arch
|
2021-03-03 22:53:02 +05:30 |
|
Mahesh Doijade
|
908dddb207
|
cudaCompressibleMemory: refactor and refine, and use only cuMemMap API for non-compressible allocation.
hence, remove use of cudaMalloc.
|
2020-05-27 19:00:33 +05:30 |
|
Mahesh Doijade
|
56b5b4c767
|
-- add cudaCompressibleMemory sample
-- fix simpleD3D12 build issue
-- fix simpleVulkanMMAP shader file locating issue
|
2020-05-22 22:07:30 +05:30 |
|