search

TiledTensor / TiledCUDA

TiledCUDA is a highly efficient kernel template library designed to elevate CUDA C’s level of abstraction for processing tiles.
MIT License
159 stars 10 forks source link

feat(kernel): Add Batched Gemm kernel. #11

Closed KuangjuX closed 7 months ago

KuangjuX commented 7 months ago

Close #6