Anthony Chang 76764d8c92 Manual control of MAC cluster for improved interwave performance (#184) ...

* manual control of MAC cluster for improved 2-wave performance

ensure setprio's order; ensure inner loop size >= local read size

synchronize when single mac cluster

* format

* use value field from ck::integral_constant

* roll out inter-wave loop scheduler to c-shuffle gemm variants

will gradually roll out to other applicable device ops when occasional reg spill is resolved

* additional comments

* format

* fix mismatch between inter-wave pipeline and interwave blockwise gemm

* address review feedback

* amend

2022-05-10 19:19:22 -05:00

..

blockwise_gemm_dlops_v2r2.hpp

Compile for gfx908 and gfx90a (#130)

2022-03-31 12:33:34 -05:00

blockwise_gemm_dlops_v2r3.hpp

Compile for gfx908 and gfx90a (#130)

2022-03-31 12:33:34 -05:00

blockwise_gemm_dlops_v3.hpp

Compile for gfx908 and gfx90a (#130)

2022-03-31 12:33:34 -05:00

blockwise_gemm_xdlops.hpp

Manual control of MAC cluster for improved interwave performance (#184)

2022-05-10 19:19:22 -05:00

blockwise_tensor_slice_transfer_v5r1.hpp

Code refactor (#175)

2022-05-09 14:57:59 -05:00

reduction_functions_blockwise.hpp

Improve Reduction kernel api (#152)

2022-04-04 20:31:44 -05:00

thread_group_tensor_slice_transfer_v4r1.hpp

Code refactor (#175)

2022-05-09 14:57:59 -05:00

thread_group_tensor_slice_transfer_v6r1.hpp

Code refactor (#175)

2022-05-09 14:57:59 -05:00

thread_group_tensor_slice_transfer_v6r2.hpp

Code refactor (#175)

2022-05-09 14:57:59 -05:00

thread_group_tensor_slice_transfer_v6r3.hpp

Code refactor (#175)

2022-05-09 14:57:59 -05:00