Skip to content

feat(minimaxm3-fp4-b200-vllm): bump vLLM nightly to 5e35a6f4, enable cutlass MSA decode / 更新 minimaxm3-fp4-b200-vllm vLLM 夜版镜像至 5e35a6f4，启用 cutlass MSA 解码后端 #2468

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
functionstackx merged 6 commits into from
Aug 4, 2026
Merged

feat(minimaxm3-fp4-b200-vllm): bump vLLM nightly to 5e35a6f4, enable cutlass MSA decode / 更新 minimaxm3-fp4-b200-vllm vLLM 夜版镜像至 5e35a6f4，启用 cutlass MSA 解码后端 #2468

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
a4e3b3a
feat(minimaxm3-fp4-b200-vllm): bump vLLM nightly to 5e35a6f4, enable …
xinli-sw Aug 3, 2026
9eb368e
chore(perf-changelog): add minimaxm3-fp4-b200-vllm nightly image bump…
xinli-sw Aug 3, 2026
1450af9
fix(minimaxm3-fp4-b200-vllm): use attention_config flag for cutlass M…
xinli-sw Aug 3, 2026
31d13ef
Merge origin/main and re-append perf-changelog entry for minimaxm3-fp…
xinli-sw Aug 3, 2026
ab38a78
feat(minimaxm3-fp4-b200-vllm): add conc-conditional chunked-prefill d…
xinli-sw Aug 3, 2026
ec005c2
Merge remote-tracking branch 'origin/main' into pr-2468-reuse-52036
functionstackx Aug 4, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions benchmarks/single_node/fixed_seq_len/minimaxm3_fp4_b200.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,6 +33,7 @@ SERVER_LOG=/workspace/server.log

export VLLM_ENGINE_READY_TIMEOUT_S=3600
export VLLM_FLOAT32_MATMUL_PRECISION=high
export VLLM_MINIMAX_M3_MSA_DECODE_BACKEND=cutlass

if [ "${DP_ATTENTION}" = "true" ]; then
PARALLEL_ARGS="--tensor-parallel-size=1 --data-parallel-size=$TP --enable-expert-parallel"
Expand Down
2 changes: 1 addition & 1 deletion configs/nvidia-master.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8303,7 +8303,7 @@ minimaxm3-fp8-b200-vllm:
# (MSA sparse/index cache); weights are pre-staged at /scratch/fsw/models/MiniMax-M3-NVFP4
# (launch_b200-dgxc.sh resolves MODEL_PATH for minimaxm3-fp4).
minimaxm3-fp4-b200-vllm:
image: vllm/vllm-openai:nightly-4080263bb2c5d10deac17aaeb88e0823bc35bca9
image: vllm/vllm-openai:nightly-5e35a6f4f9bbc217c599692157ca985c894373f7
model: nvidia/MiniMax-M3-NVFP4
model-prefix: minimaxm3
runner: b200-dgxc
Expand Down
7 changes: 7 additions & 0 deletions perf-changelog.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5355,3 +5355,10 @@
- "Apply the accuracy-gated Kimi-K2.5 MXFP4 settings: tuned AITER MXFP4 MoE, fused shared experts, FP8 KV cache, block size 16, 16384 batched tokens, 512 sequences, async scheduling, gpu-memory-utilization 0.85 (headroom for CUDA-graph capture on MI355X), and the AITER BF16 GEMM path"
- "Extend the TP4 and TP8 8k1k concurrency sweep from 64 to 128 (1k1k deprecated per #2263)"
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2213

- config-keys:
- minimaxm3-fp4-b200-vllm
description:
- "Bump vLLM nightly image to nightly-5e35a6f4f9bbc217c599692157ca985c894373f7 (fixed-len MiniMax-M3 fix)"
- "Enable VLLM_MINIMAX_M3_MSA_DECODE_BACKEND=cutlass"
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2468
Loading