composable_kernel icon indicating copy to clipboard operation
composable_kernel copied to clipboard

Kernarg latency opt

Open hsadasiv opened this issue 2 years ago • 0 comments

This depends on PR#1028. The new files modified are few:

        modified:   example/53_gemv_splitk/CMakeLists.txt
        modified:   example/54_tall_and_skinny_gemm_splitk/CMakeLists.txt
        modified:  example/54_tall_and_skinny_gemm_splitk/run_tall_and_skinny_gemm_splitk_example.inc
        modified:   include/ck/host_utility/kernel_launch.hpp
        modified:   include/ck/tensor_operation/gpu/device/impl/device_tall_and_skinny_gemm_splitk.hpp
        modified:   include/ck/tensor_operation/gpu/grid/gridwise_tall_and_skinny_gemm_splitk.hpp
       conflict resolved: library/src/tensor_operation_instance/gpu/CMakeLists.txt

hsadasiv avatar Dec 13 '23 20:12 hsadasiv