FICHA · MANJARO

onnxruntime-rocm

Cross-platform, high performance scoring engine for ML models (with ROCm)

  • Runtime
  • RUNTIME
  • MACHINE-LEARNING
  • ONNX
  • Dependency only
official+codex · reviewed · May 28, 2026 description in en

Description

Provides ONNX Runtime with ROCm support for running machine learning models on compatible AMD GPUs. It can accelerate inference when the model and hardware stack are supported.

Use it when software needs ONNX inference on AMD GPU hardware. Unsupported operators, driver issues, or insufficient VRAM can cause failures or slow execution.

Permissions

Permissions not analysed for this source yet.