Support GPU co-location for improved utilization
Open
@balajinvda is already working on this.
Since Jul 15, 2026.
roadmap
- Dominant language
- Go
- Stars
- 218
- Forks
- 72
- Avg merge
- 1d 12h
- Merged PRs (30d)
- 427
Description
Is this related to a problem? Please describe.
Some deployments reserve GPUs but use them intermittently. This can leave GPU capacity idle even when other compatible workloads could use it.
Describe the solution you'd like
Add an explicit, opt-in GPU co-location policy that allows compatible workloads to share a GPU while preserving isolation and predictable performance. The design should define and bound memory, compute, failure, and performance interference, make that interference observable, and allow co-location to be controlled per deployment or workload class.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Assessment
This issue has not been assessed yet.