kyma-project / kyma-project/gpu
GPU Module Initiative
Open
@dusglumac is already working on this.
Since May 7, 2026.
epic
- Dominant language
- Go
- Stars
- 2
- Forks
- 3
- PR merge metrics
- No merged PRs in 30d
Description
GPU Module
Goal: Deliver the Kyma GPU Module - a Kubernetes operator that automatically installs and manages the NVIDIA GPU Operator on SKR clusters.
How it works (in plain terms): A user enables the GPU module from BTP Cockpit. Kyma Lifecycle Manager creates a Gpu resource in the cluster. The operator detects GPU nodes, installs the full NVIDIA stack via Helm with the correct configuration for Garden Linux, and reports health status back. GPU workloads (e.g. AI/ML containers) can then run on the cluster.
Epics
| Epic | Description | Done |
|---|---|---|
| Project scaffold | Repo setup | [x] |
| API and CRD | Gpu resource definition, validation, generated manifests |
[x] |
| Node detection | Identify GPU nodes by machine type and OS | [x] |
| Helm install | Module enabled on a real cluster, NVIDIA stack running, smoke test passes | [ ] |
| Status aggregation | Component health rolled up into Gpu.status |
[ ] |
| Configuration and lifecycle | Spec changes, drift correction, uninstall, mixed-OS clusters | [ ] |
| Module packaging | Deployment, RBAC, ModuleTemplate, Dockerfile, KLM integration | [ ] |
| Documentation | User guide, CRD reference, troubleshooting, ops guide | [ ] |
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.