[Hardware][NVIDIA] FP4 MoE kernel optimization (#19110)

Signed-off-by: Chiyue Wei <chiyuew@nvidia.com>
Co-authored-by: Chiyue Wei <chiyuew@nvidia.com>
This commit is contained in:
Chiyue Wei
2025-06-05 09:48:26 -07:00
committed by GitHub
co-authored by Chiyue Wei
parent ec89524f50
commit 61059bee40
12 changed files with 165 additions and 38 deletions
+4 -1
View File
@@ -80,7 +80,10 @@ def test_cutlass_fp4_moe_no_graph(m: int, n: int, k: int, e: int, topk: int,
w2[expert], w2_gs[expert])
score = torch.randn((m, e), device="cuda", dtype=dtype)
topk_weights, topk_ids = fused_topk(a, score, topk, renormalize=False)
topk_weights, topk_ids, _ = fused_topk(a,
score,
topk,
renormalize=False)
a1_gs = torch.ones((e, ), device="cuda", dtype=torch.float32)
a2_gs = torch.ones((e, ), device="cuda", dtype=torch.float32)