Po Yen Chen 144377ae38 [CK_TILE] FMHA forward batch_prefill optimization for low CU utilization (#2251) ...

* Add constraint on traits/tile/pipeline

* Use kM0=128 if max_seqlen_q == 8192

* Re-format codegen script

* Remove redundant attr name postix

* Fix import error: default field in dataclass

* Use kK0=64 & kK1=64 to hide latency

* Use CU utilization to decide tile size

[ROCm/composable_kernel commit: 28cd0dffc9]

2025-05-29 18:36:33 +09:00

..

01_fmha

[CK_TILE] FMHA forward batch_prefill optimization for low CU utilization (#2251)

2025-05-29 18:36:33 +09:00

02_layernorm2d

Revert "Add ck tile examples to package (#1880)" (#2150)

2025-04-30 10:20:16 -07:00

03_gemm

Revert "[CK_tile] Add rotating buffer feature for universal gemm (#2200)" (#2256)

2025-05-28 09:46:52 -06:00

04_img2col

Revert "Add ck tile examples to package (#1880)" (#2150)

2025-04-30 10:20:16 -07:00

05_reduce

Revert "Add ck tile examples to package (#1880)" (#2150)

2025-04-30 10:20:16 -07:00

06_permute

Revert "Add ck tile examples to package (#1880)" (#2150)

2025-04-30 10:20:16 -07:00

09_topk_softmax

Revert "Add ck tile examples to package (#1880)" (#2150)

2025-04-30 10:20:16 -07:00

10_rmsnorm2d

Revert "Add ck tile examples to package (#1880)" (#2150)

2025-04-30 10:20:16 -07:00

11_add_rmsnorm2d_rdquant

Fix 11_add_rmsnorm2d_rdquant (#2207)

2025-05-20 15:15:28 -07:00

12_smoothquant

Revert "Add ck tile examples to package (#1880)" (#2150)

2025-04-30 10:20:16 -07:00

13_moe_sorting

Add catch blocks in example GEMM apps to enable better error handling (Issue: 1928) (#2234)

2025-05-27 22:32:42 -07:00

14_moe_smoothquant

Revert "Add ck tile examples to package (#1880)" (#2150)

2025-04-30 10:20:16 -07:00

15_fused_moe

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

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

16_batched_gemm

Add catch blocks in example GEMM apps to enable better error handling (Issue: 1928) (#2234)

2025-05-27 22:32:42 -07:00

17_grouped_gemm

Add catch blocks in example GEMM apps to enable better error handling (Issue: 1928) (#2234)

2025-05-27 22:32:42 -07:00

18_flatmm

Revert "[CK_tile] Add rotating buffer feature for universal gemm (#2200)" (#2256)

2025-05-28 09:46:52 -06:00

35_batched_transpose

Vectorized Transpose for Batched Transpose CK Tile Operator (#2131)

2025-05-12 00:41:45 -07:00

36_copy

Simple copy kernel, which can be a tool to experiment with CK_Tile API with minimal code. (#2156)

2025-05-07 00:02:59 -07:00

CMakeLists.txt

Simple copy kernel, which can be a tool to experiment with CK_Tile API with minimal code. (#2156)

2025-05-07 00:02:59 -07:00

remod.py

introducing ck_tile! (#1216)

2024-04-15 19:27:12 -05:00