vllm-project/vllm-omni

[RFC]: MoE performance optimizations

Open

#5,200 opened on Jul 19, 2026

View on GitHub
 (4 comments) (1 reaction) (1 assignee)Python (1,067 forks)github user discovery
Kernel optimizationRFCdiffusionhelp wanted

Repository metrics

Stars
 (4,990 stars)
PR merge metrics
 (PR metrics pending)

Description

Motivation.

Mixture-of-Experts (MoE) layers are a large fraction of end-to-end latency for MoE diffusion models served by vLLM-Omni (e.g. HunyuanImage-3.0). On our L20X / Hopper stack we profiled HunyuanImage TP=4 (tests/dfx/perf/tests/test_hunyuan_image_tp4.json workload shape: 1024×1024, 8 steps) with torch.profiler:

Therefore, optimizing MoE is critical to improving overall performance.

Proposed Change.

This RFC is a tracking issue. Concrete PRs can link back here.

(1) vLLM-Omni: fine-tuned MoE JSON support

we use ncu to profiling fused_moe_kernel, and found that sm in low utilazation:

After this, the ncu profling is:

(2) Upstream vLLM: further fused_moe optimization under TMA

On SM90+, use tl.make_tensor_descriptor (TMA) for expert weight (B) loads in fused_moe_kernel, keeping a pointer-load fallback for older arches. Goal: raise effective memory throughput toward compute-bound when configs are already reasonable.

After this, the ncu profling is:

Call for contributions

Please keep profiling / NCU notes and new ideas on this issue — Happy to co-triage and turn solid findings into follow-up PRs.

Feedback Period.

No response

CC List.

No response

Any Other Things.

No response

Before submitting a new issue...

  • Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.

Contributor guide