Amazon SageMaker HyperPod supports flexible instance groups

AI Impact Summary

Amazon SageMaker HyperPod now supports flexible instance groups, allowing customers to manage instance types and subnets within a single group. This eliminates the operational overhead of managing multiple instance groups for resilience, cost optimization, and subnet utilization. Customers can now leverage automatic priority-based fallback across instance types without manual retries, simplifying scaling and improving performance for both training and inference workloads.