[NV] H200 GLM5 fp8 update sglang container

[hshrivastava-droid]

Summary

Update the GLM-5 FP8 H200 SGLang benchmark configuration and launch script:

• SGLang image: Change from lmsysorg/sglang:v0.5.12-cu130 to lmsysorg/sglang:v0.5.11-cu129
• Runner: Switch from h200 to h200-dgxc
• Launch option: Add --enable-flashinfer-allreduce-fusion to the server launch command for improved allreduce performance

Changed Files

File	Change
.github/configs/nvidia-master.yaml	Updated image tag and runner type for glm5-fp8-h200-sglang
benchmarks/single_node/glm5_fp8_h200.sh	Added --enable-flashinfer-allreduce-fusion flag
perf-changelog.yaml	Added changelog entry for these config changes

Validation

• Validation run

[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://cookbook.sglang.io/
• https://docs.vllm.ai/projects/recipes/en/latest/

If it is not, please create a PR first before we can merge your 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/sgl-cookbook

[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://cookbook.sglang.io/
• https://docs.vllm.ai/projects/recipes/en/latest/

If it is not, please create a PR first before we can merge your 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/sgl-cookbook

[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://cookbook.sglang.io/
• https://docs.vllm.ai/projects/recipes/en/latest/

If it is not, please create a PR first before we can merge your 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/sgl-cookbook

[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://cookbook.sglang.io/
• https://docs.vllm.ai/projects/recipes/en/latest/

If it is not, please create a PR first before we can merge your 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/sgl-cookbook

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=25872417396
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=25872417396

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=25872417396
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=25872417396

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=25874676435
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=25874676435

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=25875894361
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=25875894361

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=25875894361
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=25875894361

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=25875894361
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=25875894361

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=25875894361
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=25875894361

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=26113503153
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=26113503153

[hshrivastava-droid]

@claude add PR description

[Klaud-Cold]

Claude finished @hshrivastava-droid's task in 37s —— View job

---

Adding PR Description

• Gather context and analyze changes
• Write and update PR description

Added a PR description covering the three key changes in this PR:

1. SGLang image update (v0.5.12-cu130 → v0.5.11-cu129)
2. Runner switch (h200 → h200-dgxc)
3. Added --enable-flashinfer-allreduce-fusion launch flag

Also included links to the validation run and a summary table of changed files.

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=26118912054
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=26118912054

[hshrivastava-droid]

sglang cookbook- sgl-project/sgl-cookbook#276

[hshrivastava-droid]

@functionstackx could you please help reviewing this

[functionstackx]

can u doc update to correct repo https://github.com/sgl-project/sgl-cookbook
is depreipcated

https://github.com/sgl-project/sglang/tree/main/docs_new
is the new repo

[hshrivastava-droid]

updated sglang receipe- sgl-project/sglang#25814
@functionstackx

[functionstackx]

/reuse-sweep-run

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=26130688982
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=26130688982