Manage Amazon SageMaker HyperPod Spaces directly from SageMaker Studio
Amazon SageMaker now lets data scientists create, configure, start, stop, and open SageMaker Spaces on HyperPod EKS clusters directly from the SageMaker Studio UI, removing the need for HyperPod CLI or kubectl commands. The new IDE and Notebooks tab on a HyperPod cluster detail page provides a guided form for compute, namespace, storage, and image settings, plus a searchable table of all Spaces with status and resource allocations. This UI‑first workflow shortens the time from cluster access to productive development to a few clicks, enabling interactive JupyterLab or Code Editor sessions alongside training and inference workloads on the same GPU infrastructure.
⚡ Key Takeaways
- Create Spaces with configurable compute, namespaces, storage, and image settings via a guided form in Studio.
- Space management UI shows name, application type, status, access type, storage, GPU, and vCPU allocations in a searchable table.
- Start and stop Spaces with a single selection to free compute resources when idle.
- Open Spaces directly in the browser (JupyterLab or Code Editor) or connect through a remote IDE such as VS Code.
- Administrators install the SageMaker Spaces add‑on on HyperPod EKS clusters via a one‑click “Quick install” or a “Custom install” that also configures web UI access.
Engineers can now provision and manage interactive development environments on large‑scale HyperPod clusters without leaving Studio, reducing operational overhead and improving GPU utilization for both training and inference workloads.
✅ Practical Steps
- From the Amazon SageMaker AI console, open the HyperPod EKS cluster’s IDE and Notebooks tab and use the Quick install (or Custom install) to add the SageMaker Spaces add‑on.
- In SageMaker Studio, navigate to the IDE and Notebooks tab on a HyperPod cluster detail page, then use the “Create Space” form to specify compute, namespace, storage, and image settings.
- Use the table actions to start, stop, or open a Space directly in the browser or via a remote IDE like VS Code.
Want the full story? Read the original article.
Read on AWS ML Blog ↗