What changed
Amazon SageMaker AI added ordered instance preference lists for training and processing jobs. A submission can specify alternative instance types and counts, then start on the first configuration with capacity, using on-demand capacity or reserved SageMaker Flexible Training Plans.
Why it matters
Teams no longer need parallel submissions or custom retry loops to search for scarce GPU capacity. The selected hardware can vary between runs, so reproducibility, performance and cost controls must move into job metadata and acceptance tests.
What to do
List only configurations already validated for the workload. Record the resolved instance type and count with the model, container, seed, dataset, output, runtime and cost; force a fallback in a canary and verify numerical tolerances, quotas, cleanup and cancellation before production use.
Caveat
AWS says the feature is generally available, but it does not guarantee capacity or equivalent runtime, numerics and price across alternatives. The claim that jobs will likely start sooner is provider guidance rather than a service-level commitment.