{"data":{"slug":"hami","name":"HAMi","tagline":"Apache-2.0 CNCF project that shares and slices GPUs and other accelerators across Kubernetes workloads","category":"devops","categoryName":"DevOps","url":"https://project-hami.io","github":"https://github.com/Project-HAMi/HAMi","pricing":"open-source","hook":"HAMi lets several Kubernetes pods share one GPU, each with its own memory and compute limit, so a cluster full of small inference and notebook jobs stops reserving a whole card per pod.","whatsGood":"Apache-2.0, a CNCF Incubating project, release v2.10.0. Installs with Helm as a scheduler extender plus device plugins; pods ask for a slice with resource requests such as GPU memory and core percentage. Supports NVIDIA plus several other accelerator vendors, and needs no changes to your application images. ★~4.7k.","theCatch":"It is a scheduling layer for a GPU cluster you already run, not a model server — you still bring vLLM, Triton or whatever serves the model. How limits are enforced depends on each vendor’s device backend, so isolation and overcommit behaviour differ between NVIDIA and other accelerators; test the limits on your hardware before you rely on them.","verdict":"The CNCF way to pack many small GPU jobs onto shared cards — validate isolation per accelerator vendor.","addedAt":"2026-10-12","featured":false,"domain":"AI Agents","subSpecialty":"GPU Scheduling","subcategory":"GPU Scheduling","tags":["self-host","permissive","local-llm","kubernetes","active"],"capabilities":["kubernetes","gpu"],"surfaces":["Self-host"],"ecosystem":["Go"],"licenseModel":"Apache-2.0","githubStars":4851,"starsPerWeek":18,"access":"free","archived":false,"repo":{"createdAt":"2021-09-14","pushedAt":"2026-10-19","license":"Apache-2.0","archived":false,"checkedAt":"2026-10-19"}},"related":[{"slug":"github-actions","name":"GitHub Actions","tagline":"CI/CD that lives where your code does"},{"slug":"docker","name":"Docker","tagline":"Container runtime that changed deployment"},{"slug":"kubernetes","name":"Kubernetes","tagline":"Container orchestration at scale"},{"slug":"pulumi","name":"Pulumi","tagline":"Infrastructure as code in real languages"},{"slug":"fly-io","name":"Fly.io","tagline":"Deploy apps to VMs close to your users"}],"links":{"self":"/api/gems/hami","html":"/gems/hami","markdown":"/gems/hami.md","category":"/api/gems?category=devops","categoryPage":"/stacks/devops"}}