ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-04 05:31:24 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix CK Tile DP + 2 Tile Stream-K Validation Errors (#3269)

Browse Source 
When there are multiple workgroups contributing to a tile, when using
atomics, there may be round off error in cases where the accumulator
type is not the same as the C type. To compute an error tolerance for
test validation, the Stream-K Tile Partitioner has a function called
estimate_num_wgs_per_tile to estimate the number of workgroups per tile.
That said, this function only provides an estimate. In some cases for
DP+2TSK, the function returns 1 rather than the more accurate value of
2.

Thus, this change updates the estimate_num_wgs_per_tile function to
explicitely return the value of 2 in cases for DP+2TSK to ensure that we
have a better error tolerance to avoid test failures due to round-off
error.
This commit is contained in:
Emily Martins
2025-11-21 20:29:47 -07:00
committed by GitHub
parent 21ae743acd
commit 02ab76c2cb
3 changed files with 18 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
test/ck_tile/gemm_streamk/CMakeLists.txt
View File

@@ -13,7 +13,7 @@ list(APPEND EXAMPLE_GEMM_COMPILE_COMPUTE_V4_OPTIONS
set(EXAMPLE_GEMM_COMPILE_COMPUTE_ASYNC_OPTIONS ${EXAMPLE_GEMM_COMPILE_COMPUTE_V4_OPTIONS})
# Currently test_ck_tile_streamk_smoke is only built on gfx9
if(GPU_TARGETS MATCHES "gfx9")
if(GPU_TARGETS MATCHES "gfx90a|gfx942|gfx950")
include_directories(BEFORE ${CMAKE_CURRENT_SOURCE_DIR})