carlushuang 652263d073 [CK_TILE] optimize moe sorting kernel, boost large context case up to 20x (#2153) ...

* combine 2-3 as single stage

* support zeroing

* improve long tokens

* update specialization

* b16 ws

* 8bit topk optimize

* update 15 example

[ROCm/composable_kernel commit: 4e9b76f88c]

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

..

fused_moe_api.cpp

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

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

fused_moegemm_api_internal.hpp

[CK_TILE] Fix mock token id, support g1u1/g1u0 through same inline code block (#1808)

2025-01-16 17:51:10 +08:00

fused_moegemm_api_traits.hpp

[CK_TILE] Fix mock token id, support g1u1/g1u0 through same inline code block (#1808)

2025-01-16 17:51:10 +08:00

fused_moegemm_api.cpp

[CK_TILE] Fix mock token id, support g1u1/g1u0 through same inline code block (#1808)

2025-01-16 17:51:10 +08:00

fused_moegemm_bf16_m32.cpp

[CK_TILE] Fix mock token id, support g1u1/g1u0 through same inline code block (#1808)

2025-01-16 17:51:10 +08:00

fused_moegemm_fp16_m32.cpp

[CK_TILE] Fix mock token id, support g1u1/g1u0 through same inline code block (#1808)

2025-01-16 17:51:10 +08:00

fused_moesorting_api.cpp

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

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