Skip to content

[WIP] B200 Minimax FP8 vllm upgrade #947

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

Open
kedarpotdar-nv wants to merge 10 commits into
base: main
Choose a base branch
from
+22 −10
Open

[WIP] B200 Minimax FP8 vllm upgrade #947

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
10 commits
Select commit Hold shift + click to select a range
dc27fa1
Update nvidia-master.yaml
kedarpotdar-nv Mar 26, 2026
9ee612f
vllm version bump
kedarpotdar-nv Mar 26, 2026
a27edee
add perf changelog
kedarpotdar-nv Mar 26, 2026
a33dc21
update search space and configs
kedarpotdar-nv Mar 27, 2026
ecaac19
Merge branch 'main' into nv/minimax-vllm018
kedarpotdar-nv Mar 27, 2026
89acdf2
fix typo in VLLM_USE_DEEP_GEMM
kedarpotdar-nv Mar 27, 2026
f12f56d
Merge origin/main and resolve conflicts
github-actions[bot] Mar 28, 2026
83706ae
Remove ISL 1024 / OSL 8192 seq-len config for minimaxm2.5-fp8-b200-vllm
github-actions[bot] Mar 29, 2026
e405b60
update image
kedarpotdar-nv Mar 30, 2026
e4333e9
Merge branch 'main' into nv/minimax-vllm018
kedarpotdar-nv Mar 30, 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
14 changes: 8 additions & 6 deletions .github/configs/nvidia-master.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3101,7 +3101,7 @@ gptoss-fp4-b200-vllm:
- { tp: 8, conc-start: 4, conc-end: 4 }

minimaxm2.5-fp8-b200-vllm:
image: vllm/vllm-openai:v0.17.0-cu130
image: vllm/vllm-openai:v0.18.0-cu130
model: MiniMaxAI/MiniMax-M2.5
model-prefix: minimaxm2.5
runner: b200
Expand All @@ -3112,13 +3112,15 @@ minimaxm2.5-fp8-b200-vllm:
- isl: 1024
osl: 1024
search-space:
- { tp: 2, conc-start: 4, conc-end: 64 }
- { tp: 4, conc-start: 4, conc-end: 64 }
- { tp: 2, conc-start: 4, conc-end: 256 }
- { tp: 4, conc-start: 4, conc-end: 256 }
- { tp: 4, ep: 4, conc-start: 64, conc-end: 64 }
- { tp: 8, conc-start: 4, conc-end: 8 }
- isl: 8192
osl: 1024
search-space:
- { tp: 2, conc-start: 4, conc-end: 64 }
- { tp: 4, conc-start: 4, conc-end: 64 }
- { tp: 2, conc-start: 4, conc-end: 256 }
- { tp: 4, conc-start: 4, conc-end: 256 }

gptoss-fp4-h100-vllm:
image: vllm/vllm-openai:v0.18.0
Expand Down Expand Up @@ -3345,7 +3347,7 @@ gptoss-fp4-h200-vllm:
- { tp: 8, conc-start: 4, conc-end: 32 }

minimaxm2.5-fp8-h200-vllm:
image: vllm/vllm-openai:v0.18.0
image: vllm/vllm-openai:nightly-5b8c30d62b754b575e043ce2fc0dcbf8a64f6306
model: MiniMaxAI/MiniMax-M2.5
model-prefix: minimaxm2.5
runner: h200
Expand Down
7 changes: 4 additions & 3 deletions benchmarks/single_node/minimaxm2.5_fp8_b200.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,8 +24,8 @@ hf download "$MODEL"
SERVER_LOG=/workspace/server.log
PORT=${PORT:-8888}

export VLLM_USE_FLASHINFER_MOE_FP8=0
export VLLM_MOE_USE_DEEP_GEMM=0
export VLLM_USE_DEEP_GEMM=0
export VLLM_FLASHINFER_ALLREDUCE_BACKEND=mnnvl

if [ "$EP_SIZE" -ge 1 ]; then
EP=" --enable-expert-parallel"
Expand All @@ -47,7 +47,8 @@ $EP \
--gpu-memory-utilization 0.95 \
--max-model-len $MAX_MODEL_LEN \
--block-size=32 \
--no-enable-prefix-caching \
--kv-cache-dtype fp8 \
--stream-interval 20 --no-enable-prefix-caching \
--trust-remote-code > $SERVER_LOG 2>&1 &

SERVER_PID=$!
Expand Down
11 changes: 10 additions & 1 deletion perf-changelog.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1136,7 +1136,7 @@
description:
- "Disable prefix caching (--no-enable-prefix-caching) for all MiniMax benchmarks using random datasets"
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/966

- config-keys:
# NVIDIA single-node
- dsr1-fp4-b200-sglang
Expand Down Expand Up @@ -1213,3 +1213,12 @@
- "Uses nvidia/GLM-5-NVFP4 model with modelopt_fp4 quantization"
- "Image: lmsysorg/sglang:nightly-dev-cu13-20260328-a27651d5"
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/973

- config-keys:
- minimaxm2.5-fp8-b200-vllm
description:
- "Update vLLM image from v0.17.0 to v0.18.0 for MiniMax-M2.5 FP8 B200"
- "Add tp4 ep4 search-space entries (conc 32-256) for all seq-len configs"
- "Remove ISL 1024 / OSL 8192 seq-len config"
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/947

Loading