[CI] Flash Attention RDNA CI

[micmelesse]

Motivation

The triton backend of flash attention is used on consumer cards (RDNA) in addition to datacenter cards (CDNA). This adds an RDNA CI runner (aiter-gfx1100) to the existing flash_attention_integration workflow via matrix strategy so both architectures are tested in parallel. This is a follow up to #1974

Technical Details

Test Plan

Test Result

Submission Checklist

• Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

[github-actions]

🏷️ CI Guide

Runs automatically on every PR:

• ✅ Pre-checks (submodule verification, code formatting)
• ✅ Aiter op tests (gfx942 + gfx950)
• ✅ Triton tests (only when aiter/ops/triton/** or related paths are changed)

Extended tests (opt-in via labels):

Label	Tests
ci:sglang	SGLang integration tests
ci:atom	ATOM benchmark (DeepSeek-R1 + GPT-OSS)
ci:multi-gpu	Multi-GPU op tests (8 GPU)
ci:vllm	vLLM benchmark
ci:all	All of the above

Add labels via the sidebar or gh pr edit 2222 --add-label <label>

[Copilot]

Pull request overview

Adds RDNA (consumer GPU) coverage to the existing Flash Attention Triton integration workflow by running the same job on both CDNA (MI355) and RDNA3 runners in parallel, improving CI validation for the Triton backend across supported AMD architectures.

Changes:

• Convert the Triton integration job to a matrix strategy to run on both MI355 and RDNA3 runners.
• Make job names and benchmark artifacts/logs architecture-specific to avoid collisions across matrix runs.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

You can also share your feedback on Copilot code review. Take the survey.