Manage Amazon SageMaker HyperPod Spaces Directly from SageMaker Studio
Developers and data scientists can now manage Amazon SageMaker HyperPod Spaces entirely from the SageMaker Studio user interface, eliminating the prior reliance on CLI tools. This speeds up environment creation and access for machine learning development.
What changed?
Amazon Web Services has introduced a visual management interface for SageMaker HyperPod Spaces within the SageMaker Studio UI. Data scientists and ML engineers can now create, configure, start, stop, and open Spaces directly from their browser. JupyterLab and Code Editor environments are accessible in a few clicks, and the new interface provides a guided form for configuring resources and governance. The update removes the need to use the HyperPod CLI or kubectl for routine Space operations.
Why does it matter to an everyday developer?
This update allows ML practitioners and developers to launch and manage development environments on high-performance GPU clusters without learning or maintaining CLI workflows. Spaces can be managed from a single UI, speeding up access to Jupyter and VS Code-style environments for model training and experimentation. Developers can also view all Spaces in a searchable table with status, compute allocations, and access controls, making cluster resource management more transparent and accessible.
What can the developer do now?
Once the Spaces add-on is enabled by an administrator on a HyperPod EKS cluster, developers can use SageMaker Studio to create, configure, and open Spaces with desired compute, storage, and governance settings. They can choose JupyterLab or Code Editor (VS Code-like) environments in the browser, or use the AWS Toolkit in local VS Code for remote coding. Stopping and starting Spaces from the UI helps optimize resource usage. No CLI or manual Kubernetes commands are required for daily operations.