Po Yen Chen b87e256c42 [CK_TILE][FMHA] Enable dwordx4 loading in async_load_tile_raw() (#2549) ...

* Support async load dwordx4

* Enlarge load size on gfx950

[ROCm/composable_kernel commit: 4a7ecce096]

2025-08-22 10:13:47 +08:00

..

amd_buffer_addressing_builtins.hpp

[CK_TILE][FMHA] Enable dwordx4 loading in async_load_tile_raw() (#2549)

2025-08-22 10:13:47 +08:00

amd_buffer_addressing.hpp

[CK_TILE][FMHA] Enable dwordx4 loading in async_load_tile_raw() (#2549)

2025-08-22 10:13:47 +08:00

amd_transpose_load_encoding.hpp

[CK_TILE] Use read_tr in universal gemm (#2436)

2025-07-16 23:56:22 -07:00

arch.hpp

Support Wave32 in CK_TILE - Part 1 (#2594)

2025-08-18 10:08:31 -07:00

generic_memory_space_atomic.hpp

[CK_TILE] CK_TILE GEMM WMMA Support for GFX11/GFX12 (#2466)

2025-08-15 16:22:27 -07:00

utility.hpp

Re-enable optimization for gfx950 fmha fwd (#2671)

2025-08-13 14:57:43 +08:00

workgroup_barrier.hpp

[CK_TILE] optimize moe sorting kernel, boost large context case up to 20x (#2153)

2025-05-06 17:32:07 +08:00