GPUStack Runtime

GPUStack Runtime offers a unified interface for detecting GPU resources and managing GPU workloads.

Features

• Detect a wide range of GPU and accelerator resources:

  • AMD GPU
  • Ascend NPU
  • Cambricon MLU
  • Hygon DCU
  • Iluvatar GPU
  • MetaX GPU
  • Moore Threads GPU
  • NVIDIA GPU
  • T-Head PPU

• Manage GPU workloads on the following platforms:

  • Docker
  • Kubernetes
  • Podman (>=4.9, experimental support via the CONTAINER_HOST=http+unix:///path/to/podman/socket environment variable)

Contributions to support additional GPU resources are welcome!

Installation

pip install gpustack-runtime

Documentation

The API reference renders from docstrings. Build the site locally with make docs and open site/index.html.

Contributing

Contributions are welcome. Commits must carry a Signed-off-by line certifying the Developer Certificate of Origin.

Using GPUStack Runtime in production? Add your company or project to ADOPTERS.md via pull request.

License

Copyright (c) 2026 The GPUStack Authors. Licensed under the Apache License 2.0.