ViperEkura 2c0b5d0b5e perf: enable MMA decode path for G=1 full attention ...

- inline decode_use_mma() into dispatch_decode()
- drop G>1 guard, MMA works correctly for G>=1
- decode is memory-bound, tensor cores + cp.async still win at G=1

2026-07-11 11:45:13 +08:00

..

attn_common.h

refactor: template AttentionParams, rename .cuh to .h

2026-07-11 11:03:14 +08:00

attn_decode_split_kv_mma.cuh

refactor: template AttentionParams, rename .cuh to .h

2026-07-11 11:03:14 +08:00

attn_decode_split_kv.cuh

refactor: template AttentionParams, rename .cuh to .h

2026-07-11 11:03:14 +08:00

attn_decode.cu

perf: enable MMA decode path for G=1 full attention

2026-07-11 11:45:13 +08:00

attn_entry_utils.cuh

refactor: template AttentionParams, rename .cuh to .h

2026-07-11 11:03:14 +08:00

attn_mma_utils.cuh

refactor: template AttentionParams, rename .cuh to .h

2026-07-11 11:03:14 +08:00

attn_prefill_split_q_mma.cuh

refactor: template AttentionParams, rename .cuh to .h

2026-07-11 11:03:14 +08:00

attn_prefill_split_q.cuh

refactor: template AttentionParams, rename .cuh to .h

2026-07-11 11:03:14 +08:00

attn_prefill.cu

refactor: template AttentionParams, rename .cuh to .h

2026-07-11 11:03:14 +08:00