SageMaker Now Auto-Selects GPUs From Priority Lists

Amazon SageMaker AI now supports instance preference lists for training and processing jobs, letting users specify up to five GPU instance types in priority order. The platform automatically selects the first available option, eliminating manual retry loops and reducing wait times during peak demand. The feature addresses a common pain point where jobs pinned to a single instance type stall due to insufficient capacity. It also integrates with reserved capacity options like Flexible Training Plans, allowing teams to try reservations first before falling back to on-demand alternatives automatically.