Amazon SageMaker HyperPod Enhances Machine Learning Infrastructure with Scalability and Customizability.

Amazon SageMaker HyperPod is a scalable infrastructure for machine learning that optimizes foundation model training and inference at scale. It reduces training time by up to 40% and offers persistent clusters with built-in resiliency. SageMaker HyperPod supports Amazon Elastic Kubernetes Service (Amazon EKS) and offers two new features: continuous provisioning and custom AMIs, enhancing control and flexibility for production deployment of large-scale ML workloads.