HAMi Workshop

HAMi Workshop is a community learning resource for HAMi, the GPU virtualization middleware for Kubernetes.

Available in English and 中文.

What You'll Learn

• How the GPU software stack works, from hardware to Kubernetes scheduling
• Installing HAMi on a real GPU cluster or a local fake GPU environment
• GPU sharing, memory isolation, and fine-grained scheduling
• Monitoring GPU utilization across your cluster

Quick Start

# Install dependencies
make install

# Preview locally at http://127.0.0.1:8000
make serve

# Build static site
make build

Requires Python 3.12+.

Repository Structure

docs/
  concepts/        Background knowledge (GPU stack, drivers, HAMi architecture)
  labs/             Step-by-step lab instructions
  images/           Screenshots
  stylesheets/      Theme overrides
  javascripts/      Mermaid diagram rendering and zoom
examples/           Lab code, manifests, and configs (numbered by lab)
mkdocs.yml          Site configuration, i18n, navigation

Contributing

1. Docs live in docs/. English is the default (page.md), Chinese uses the .zh.md suffix.
2. Both language versions must exist for every page listed in nav.
3. Code and manifests go in examples/, numbered by lab.
4. Mermaid diagrams use %% title: comments for captions.
5. Run make build before pushing to verify there are no broken links.

License

Apache-2.0