Amazon SageMaker AI cuts generative AI inference scale-out time by up to half with container image caching

Pre-caching container images during scale-out events eliminates ECR pull latency, enabling up to 2x faster endpoint scaling for large generative AI models.