Modify instances.

* set KPerBlock to 64 * maximize wherever possible vector load size.

Modify instances.
* set KPerBlock to 64 * maximize wherever possible vector load size.
e8b53655 · Adam Osewski · cdec576c · e8b53655 · e8b53655 · e8b53655
Commit e8b53655 authored Aug 10, 2023 by Adam Osewski
4 changed files
--- a/library/src/tensor_operation_instance/gpu/grouped_gemm/device_grouped_gemm_xdl_splitk_f16_f16_f16_mk_kn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/grouped_gemm/device_grouped_gemm_xdl_splitk_f16_f16_f16_mk_kn_mn_instance.cpp
@@ -6,8 +6,7 @@
 #include "ck/ck.hpp"
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
 #include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
-// #include "ck/tensor_operation/gpu/device/impl/device_grouped_gemm_xdl_splitk_cshuffle.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_grouped_gemm_xdl_splitk_cshuffle.hpp"
-#include "ck/tensor_operation/gpu/device/impl/device_grouped_gemm_xdl_splitk_cshuffle_tile_loop.hpp"
 #include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"

--- a/library/src/tensor_operation_instance/gpu/grouped_gemm/device_grouped_gemm_xdl_splitk_f16_f16_f16_mk_kn_mn_irregular_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/grouped_gemm/device_grouped_gemm_xdl_splitk_f16_f16_f16_mk_kn_mn_irregular_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/grouped_gemm/device_grouped_gemm_xdl_splitk_f16_f16_f16_mk_nk_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/grouped_gemm/device_grouped_gemm_xdl_splitk_f16_f16_f16_mk_nk_mn_instance.cpp
@@ -6,8 +6,7 @@
 #include "ck/ck.hpp"
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
 #include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
-// #include "ck/tensor_operation/gpu/device/impl/device_grouped_gemm_xdl_splitk_cshuffle.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_grouped_gemm_xdl_splitk_cshuffle.hpp"
-#include "ck/tensor_operation/gpu/device/impl/device_grouped_gemm_xdl_splitk_cshuffle_tile_loop.hpp"
 #include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"

--- a/library/src/tensor_operation_instance/gpu/grouped_gemm/device_grouped_gemm_xdl_splitk_f16_f16_f16_mk_nk_mn_irregular_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/grouped_gemm/device_grouped_gemm_xdl_splitk_f16_f16_f16_mk_nk_mn_irregular_instance.cpp