[ROCm][Perf][DSV4] Improve sparse decode reduction occupancy on gfx950 (#48788)

Signed-off-by: fai <fangzhouai@gmail.com>
2026-07-17 10:24:59 -07:00
parent 11d291511a
commit efed8a1e83
1 changed files with 2 additions and 2 deletions
@@ -2205,7 +2205,7 @@ def _rocm_sparse_attn_decode_ragged_triton(
        num_warps=4,
    )

-    _sparse_attn_decode_reduce_kernel[(num_queries, heads_blocks)](
+    _sparse_attn_decode_reduce_kernel[(num_queries, num_heads)](
        part_m,
        part_l,
        part_acc,
@@ -2221,7 +2221,7 @@ def _rocm_sparse_attn_decode_ragged_triton(
        num_heads,
        HAS_ATTN_SINK=has_attn_sink,
        COMB_DIM=comb_dim,
-        BLOCK_H=block_h,
+        BLOCK_H=1,
        NUM_SPLITS=num_splits,
        SPLITS_PAD=triton.next_power_of_2(num_splits),
        num_warps=4,