Add contraction_fp64 example (#570)

* add contraction_bilinear

* add contraction_scale_xdl_fp64

* reduce tile size to avoid register spill

---------

Co-authored-by: root <root@ctr-ubbsmc16.amd.com>

[ROCm/composable_kernel commit: 24c9ee1d22]
This commit is contained in:
zjing14
2023-02-15 12:00:58 -06:00
committed by GitHub
parent 9d20a2b6b5
commit 7335ebded0
6 changed files with 854 additions and 2 deletions

View File

@@ -10,8 +10,8 @@ cmake
-D CMAKE_CXX_COMPILER=/opt/rocm/bin/hipcc \
-D CMAKE_CXX_FLAGS="-O3 -ftemplate-backtrace-limit=0 -gline-tables-only -save-temps=$PWD" \
-D CMAKE_BUILD_TYPE=Release \
-D BUILD_DEV=ON \
-D GPU_TARGETS="gfx908;gfx90a" \
-D BUILD_DEV=OFF \
-D GPU_TARGETS="gfx90a" \
-D CMAKE_VERBOSE_MAKEFILE:BOOL=ON \
-D USE_BITINT_EXTENSION_INT4=OFF \
${MY_PROJECT_SOURCE}