mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-06-11 00:39:02 +00:00
[CK_Tile] Add scale16 Support for F4 WMMA in CK_Tile ## Motivation This PR adds CK Tile support for the scale16 F4 WMMA path on gfx1250 and improves warp GEMM unit test coverage/structure for gfx1250-specific cases. ## Technical Details - Scale16 support in warp GEMM dispatch and WMMA trait plumbing: added IsScale16 plumbing to warp GEMM dispatcher path - Warp GEMM test restructuring for gfx1250: added Warp GEMM gfx1250 coverage to verify all F4 WMMA paths ## Test Plan Run ./test_ck_tile_wg_32x16x128_fp4. ## Test Result ``` ./test_ck_tile_wg_32x16x128_fp4 [----------] Global test environment tear-down [==========] 3 tests from 1 test suite ran. (1751 ms total) [ PASSED ] 3 tests. ``` ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.
ck_tile/core
ck_tile/core contains every basic functions and structures to create a GPU kernel using ck_tile. User should only include ck_tile/core.hpp this single header to use all the functionality. Everything is under ck_tile namespace. The coding style under this folder should be similar to std (snake_case for structure/function, Camel for template types...)
algorithm/
coordinate transform and some other reusable algorithm
arch/
contains some basic device building block like mma, buffer addressing, etc...
container/
contains basic container data structure, array/sequence/tuple/...
numeric/
data type, and data type related math
tensor/
tensor descriptors and tile level API
utility/
other utility function for both host/device