Issue Station
Sign In
Tools hereToolbox

vllm - 💡(How to fix) Fix [Bug]: MiniMax M2.5 TP8EP8 gfx950 with AITER causes memory access fault

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…

Error Message

See error in comment below

Fix Action

Fix / Workaround

I was able to "resolve" the issue by temporarily setting --max-num-batched-tokens 4K, so I suspect this is an issue with the M=8k shape in AITER fmoe kernel. Alternatively it could be resolved by VLLM_ROCM_USE_AITER_MOE=0.

Code Example

Your output of `python collect_env.py` here

---

docker run \
  --device=/dev/kfd \
  --device=/dev/dri   \
  --security-opt seccomp=unconfined \
  --group-add video   \
  --privileged \
  --ipc=host -p 8000:8000   \
  -v /it-share/data/:/root/.cache/huggingface \
  -e HF_HUB_OFFLINE=1  \
  -e VLLM_ROCM_USE_AITER=1 \  
  -e VLLM_ROCM_SHUFFLE_KV_CACHE_LAYOUT=1 \
  vllm/vllm-openai-rocm:latest \
  MiniMaxAI/MiniMax-M2.5 \
  --trust-remote-code \
  --tensor-parallel-size 8 \
  --enable-expert-parallel
RAW_BUFFERClick to expand / collapse

Your current environment

<details> <summary>The output of <code>python collect_env.py</code></summary> 
Your output of `python collect_env.py` here
</details>
  • vllm/vllm-openai-rocm:v0.21.0
  • 8xMI355 node

🐛 Describe the bug

Running MiniMax M2.5 on MI355X with TP8EP8 causes memory fault during startup:

docker run \
  --device=/dev/kfd \
  --device=/dev/dri   \
  --security-opt seccomp=unconfined \
  --group-add video   \
  --privileged \
  --ipc=host -p 8000:8000   \
  -v /it-share/data/:/root/.cache/huggingface \
  -e HF_HUB_OFFLINE=1  \
  -e VLLM_ROCM_USE_AITER=1 \  
  -e VLLM_ROCM_SHUFFLE_KV_CACHE_LAYOUT=1 \
  vllm/vllm-openai-rocm:latest \
  MiniMaxAI/MiniMax-M2.5 \
  --trust-remote-code \
  --tensor-parallel-size 8 \
  --enable-expert-parallel

I was able to "resolve" the issue by temporarily setting --max-num-batched-tokens 4K, so I suspect this is an issue with the M=8k shape in AITER fmoe kernel. Alternatively it could be resolved by VLLM_ROCM_USE_AITER_MOE=0.

See error in comment below

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.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING

vllm - 💡(How to fix) Fix [Bug]: MiniMax M2.5 TP8EP8 gfx950 with AITER causes memory access fault