[AMD] dsv4-fp4-mi355x-atom: enable DPA TBO at high concurrency, update image to atom0.1.4

[seungrokj]

Summary

• Update image to rocm/atom:rocm7.2.4_ubuntu24.04_py3.12_pytorch_release_2.10.0_atom0.1.4_20260612
• Enable --enable-tbo (Token-Bucket Overlap) on top of DPA+TP8 at high concurrency:
  • ISL=1024 / OSL=1024 → TBO at CONC ≥ 1024
  • ISL=8192 / OSL=1024 → TBO at CONC ≥ 256
• Update ISL=8192 search-space: TP8-only conc=4–64, DPA conc=128–1024 (previously conc=1–64 and DPA conc=64–512)
• Add perf-changelog entry

Motivation

Based on Pareto frontier analysis from ATOM runs 27367309656 and 27030375093, DPA TBO dominates the Pareto frontier at high concurrency for both ISL settings, delivering the best throughput/GPU without sacrificing interactivity vs plain DPA.

Test plan

• Verify bash -n syntax check passes on dsv4_fp4_mi355x_atom.sh
• Confirm TBO is enabled for correct ISL/CONC combinations in CI run

🤖 Generated with Claude Code

---

Note

Medium Risk
Changes benchmark CI topology and ATOM serve flags for a large model at very high concurrency, which can affect sweep stability and GPU memory even though scope is limited to perf configs.

Overview
Updates dsv4-fp4-mi355x-atom to atom0.1.4 and widens the fixed-seq-len concurrency sweeps so high-load points use DPA (and DPA+TBO where the benchmark script applies it).

In dsv4_fp4_mi355x_atom.sh, DPA+TP8 now adds --enable-tbo at CONC ≥ 1024 for 1k/1k and CONC ≥ 256 for 8k/1k. The ATOM server launch gains --no-enable_prefix_caching, --max-model-len (with eval-only context via compute_eval_context_length), matching the expanded YAML search space.

perf-changelog.yaml documents the image bump, TBO gates, and 8k/1k sweep changes.

^{Reviewed by Cursor Bugbot for commit 3f3560b. 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]

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=27416363770
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=27416363770

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=27417198968
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=27417198968

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=27418122192
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=27418122192

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=27418122192
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=27418122192