[WIP] [NV] Update MiniMax B200/B300 aggregate vLLM settings

[jasonlizhengjian]

Updates MiniMax B200/B300 aggregate vLLM image/settings.

---

Note

Low Risk
Benchmark-only image and environment tweaks for MiniMax aggregate configs; no application auth or production serving paths.

Overview
Aligns MiniMax-M2.5 aggregate vLLM benchmarks on B200/B300 (FP8 and FP4) with vLLM v0.22.0 on B300 in nvidia-master.yaml (from v0.21.0) and documents the rollout in perf-changelog.yaml.

All four fixed-seq-len launch scripts (minimaxm2.5_fp{4,8}_b{200,300}.sh) now set VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0 before serve, matching other NVIDIA recipes that disable v0.21+ CUDA-graph memory profiling to avoid startup OOM / inflated KV reservations.

^{Reviewed by Cursor Bugbot for commit c1934cb. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

Thanks for the contribution! For vLLM & SGLang, please ensure that your recipes is similar to the official vLLM recipes and/or the SGLang cookbook

• https://docs.sglang.io/cookbook/intro
• https://recipes.vllm.ai/

If it is not, please create a PR first before we can merge your single node PR into the master branch. Let's ensure that the documentation is first class such that the entire ML community can benefit from your hard work! Thank you

• https://github.com/vllm-project/recipes
• https://github.com/sgl-project/sglang/tree/main/docs_new

PR authors are responsible for ensuring that after merging, all GitHub Action jobs fully pass. A lot of the time, failures are just flakes and simply re-running the failed jobs will fix it. If re-running failed jobs is attempted, PR authors are responsible for ensuring it passes. See GitHub's docs on re-running failed jobs: https://docs.github.com/en/actions/how-tos/manage-workflow-runs/re-run-workflows-and-jobs#re-running-failed-jobs-in-a-workflow

As a rule of thumb, generally, PR authors should request a review & get a PR approval from the respective companies' CODEOWNERS before requesting a review from core maintainers.

If additional help is needed, PR authors can reach out to core maintainers over Slack.

[github-actions]

Thanks for the contribution! For vLLM & SGLang, please ensure that your recipes is similar to the official vLLM recipes and/or the SGLang cookbook

• https://docs.sglang.io/cookbook/intro
• https://recipes.vllm.ai/

If it is not, please create a PR first before we can merge your single node PR into the master branch. Let's ensure that the documentation is first class such that the entire ML community can benefit from your hard work! Thank you

• https://github.com/vllm-project/recipes
• https://github.com/sgl-project/sglang/tree/main/docs_new

PR authors are responsible for ensuring that after merging, all GitHub Action jobs fully pass. A lot of the time, failures are just flakes and simply re-running the failed jobs will fix it. If re-running failed jobs is attempted, PR authors are responsible for ensuring it passes. See GitHub's docs on re-running failed jobs: https://docs.github.com/en/actions/how-tos/manage-workflow-runs/re-run-workflows-and-jobs#re-running-failed-jobs-in-a-workflow

As a rule of thumb, generally, PR authors should request a review & get a PR approval from the respective companies' CODEOWNERS before requesting a review from core maintainers.

If additional help is needed, PR authors can reach out to core maintainers over Slack.

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=27300387564
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=27300387564