Amazon EKS support in Amazon SageMaker HyperPod to scale foundation model development

The article announces the general availability of Amazon EKS support in SageMaker HyperPod, enabling customers to run and manage their Kubernetes workloads on SageMaker HyperPod, a purpose-built infrastructure for foundation model development that reduces time to train models by up to 40%.

Specifically, the article covers:

• Integration of Amazon EKS, a managed Kubernetes service, with SageMaker HyperPod, providing benefits like automated hardware failure management and containerization capabilities.
• Features like deep health checks during cluster creation, automatic node replacement, and training resumption from checkpoints.
• Flexibility to use the HyperPod CLI or preferred tools to manage workloads, and persistent cluster environment with customization and SSM access.
• Observability through integration with CloudWatch Container Insights for monitoring node health and visualizing dashboards.
• General availability in AWS Regions where SageMaker HyperPod is available, except Europe (London).
• Links to resources like the webpage, AWS News Blog, documentation, and GitHub repository for more information.