carlushuang c12fb0a624 [CK_TILE][HOTFIX] WA for address space by disable it completely (#1947) ...

* port all moe changes from ck_moe_gemm branch

* refine codes in the pr

* fix tail odd

* fix clang format

* fix clang format2

* make hot loop scheduler compatible with 16x16 and 32x32

* clang format

* fix per token quant

* rename moe example

* clang format

* WA for address space by disable it completely

* hot fix moe gemm2

---------

Co-authored-by: coderfeli <coderfeli@163.com>
Co-authored-by: feli <felix.li@amd.com>

2025-03-06 12:01:25 +08:00

..

block

ck moe gemm implement (#1936)

2025-03-05 15:56:55 +08:00

device

Replace buffer load/store intrinsics with builtins (#1876)

2025-03-05 14:33:28 -08:00

element

Rebase the PR #1520 to ROCm repo. (#1574)

2025-02-20 18:58:14 -08:00

grid

[CK_TILE][HOTFIX] WA for address space by disable it completely (#1947)

2025-03-06 12:01:25 +08:00

thread

ck moe gemm implement (#1936)

2025-03-05 15:56:55 +08:00

warp

Remove CK_USE_AMD_MFMA_GFX950 (#1935)

2025-03-04 10:32:25 -08:00