Manage Amazon SageMaker HyperPod Spaces directly from SageMaker Studio

This article announces the ability to manage Amazon SageMaker Spaces on HyperPod EKS clusters directly from SageMaker Studio UI, eliminating the need for command-line tools.

• Create, configure, start, stop, and open Spaces through a guided form in SageMaker Studio
• Launch JupyterLab or Code Editor environments in the browser or connect via remote VS Code
• View all Spaces in a searchable table showing status, compute allocation, and quick actions
• Administrators install the Spaces add-on and configure EKS access; data scientists manage Spaces via UI
• Optional capabilities include Space templates, Task Governance quotas, Karpenter autoscaling, and idle shutdown
• Node over-provisioning reduces Space startup time from 5-7 minutes to 30-40 seconds
• No additional charges for the add-on; pay only for underlying HyperPod cluster compute

This integration streamlines access to high-performance compute infrastructure for data scientists while maintaining cost control and resource fairness through self-service cluster management.