Share GPU clusters across teams with isolation and fairness using Amazon SageMaker HyperPod

What happened

Share GPU (a chip built for many calculations at once, used for graphics and AI) clusters across teams with isolation and fairness using Amazon SageMaker (Python library) HyperPod, AWS Machine Learning Blog announced. A reference architecture for securely sharing one Amazon SageMaker HyperPod EKS cluster across multiple teams, using AWS IAM Identity Center for authentication, per-team SageMaker Domains and Kubernetes (software that runs and manages applications across many servers) namespaces for isolation, HyperPod Task Governance for fairness, and namespace-level cost allocation for chargeback. Multiple teams within the same company increasingly need shared access to expensive GPU clusters for their generative AI operations, while maintaining isolation boundaries, resource fairness, and operational independence.

Without a well-designed multi-tenant (multi-team) architecture, organizations face uncontrolled resource consumption, weak isolation between teams, an inability to attribute shared GPU costs to the teams that incur them, and administrative overhead that slows down innovation. Amazon SageMaker HyperPod is a purpose-built AI service that simplifies the management of large-scale compute clusters for gen AI workloads.