Skip to content

feat(env_config): fix ROCm GPU detection for ROCm 7.2+#139

Merged
solderzzc merged 1 commit intodevelopfrom
feature/rocm-gpu-detection
Mar 8, 2026
Merged

feat(env_config): fix ROCm GPU detection for ROCm 7.2+#139
solderzzc merged 1 commit intodevelopfrom
feature/rocm-gpu-detection

Conversation

@solderzzc
Copy link
Member

  • Add amd-smi static --json as primary detection strategy (ROCm 6.3+/7.x)
  • Keep rocm-smi as fallback for legacy ROCm <6.3
  • Add multi-GPU selection: pick GPU with most VRAM, set HIP_VISIBLE_DEVICES
  • Add _check_rocm_runtime() to verify ROCmExecutionProvider in onnxruntime (prevents CPU-only onnxruntime from shadowing onnxruntime-rocm)
  • Update deploy.sh heuristic to also check for amd-smi
  • Document GPU package conflict in requirements.txt
  • Add 14 unit tests for all ROCm detection paths

- Add amd-smi static --json as primary detection strategy (ROCm 6.3+/7.x)
- Keep rocm-smi as fallback for legacy ROCm <6.3
- Add multi-GPU selection: pick GPU with most VRAM, set HIP_VISIBLE_DEVICES
- Add _check_rocm_runtime() to verify ROCmExecutionProvider in onnxruntime
  (prevents CPU-only onnxruntime from shadowing onnxruntime-rocm)
- Update deploy.sh heuristic to also check for amd-smi
- Document GPU package conflict in requirements.txt
- Add 14 unit tests for all ROCm detection paths
@solderzzc solderzzc merged commit a6798d1 into develop Mar 8, 2026
1 check passed
@solderzzc solderzzc deleted the feature/rocm-gpu-detection branch March 8, 2026 22:30
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant