[Klaud Cold] Update dsr1-fp8-h200-trt (+mtp) TRT-LLM image to v1.3.0rc14

[functionstackx]

Summary

Update TensorRT-LLM image from v1.1.0rc2.post2 (154d/124d old) to v1.3.0rc14 (latest pre-release)

• dsr1-fp8-h200-trt: nvcr.io#nvidia/tensorrt-llm/release:1.1.0rc2.post2 → nvcr.io#nvidia/tensorrt-llm/release:1.3.0rc14
• dsr1-fp8-h200-trt-mtp: nvcr.io#nvidia/tensorrt-llm/release:1.1.0rc2.post2 → nvcr.io#nvidia/tensorrt-llm/release:1.3.0rc14

Test plan

• full-sweep-enabled sweep passes.

🤖 Generated with Claude Code

[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=26016866059
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=26016866059

[claude]

LGTM — straightforward image version bump with matching changelog entry.

Extended reasoning...

Overview

This PR bumps the TensorRT-LLM image tag from 1.1.0rc2.post2 to 1.3.0rc14 for two benchmark configs (dsr1-fp8-h200-trt and dsr1-fp8-h200-trt-mtp) in .github/configs/nvidia-master.yaml, and adds a corresponding entry in perf-changelog.yaml. No logic changes.

Security risks

None. This is a pinned image tag change for a benchmark sweep, sourced from nvcr.io/nvidia/tensorrt-llm/release — same registry/repo, just a newer pre-release tag.

Level of scrutiny

Low. The change is mechanical and follows the exact same pattern as recent image-bump PRs (#1426, #1395, #1448 in recent commits). The PR carries the full-sweep-enabled label, so the actual functional verification happens via the benchmark sweep CI.

Other factors

No bugs were flagged by the bug hunting system, no outstanding reviewer comments, and the changelog entry correctly cross-references the PR. Safe to shadow-approve.

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=26016868638
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=26016868638

[functionstackx]

Triggered re-run of the failed jobs — the underlying enroot disk issue on h200-nb_0 / h200-nb_1 is being tracked separately in #1496. The h200 SLURM partition tag has been removed from both bad runners, so the re-run should land on the healthy runners only.

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=26016868638
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=26016868638

[functionstackx]

/reuse-sweep-run

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=26045843625
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=26045843625