Amazon SageMaker HyperPod began supporting NVIDIA-powered G7e instances, which deliver substantially improved inference performance and large GPU memory capacity, enabling deployment of larger or multiple AI models simultaneously. These instances target foundation models and multimodal AI workloads, offering up to 2.3 times better throughput compared to previous generations.
Additionally, the service added r5d.16xlarge CPU instances, optimized for memory-intensive tasks like distributed data preprocessing and feature engineering. These extend HyperPod's capacity to efficiently manage both GPU and CPU heavy operations, reinforcing its capability as an all-encompassing AI development environment available across multiple AWS regions.
Source & further reading
Amazon SageMaker HyperPod now supports G7e and r5d.16xlarge instances ↗Source published 2026-04-27. This report was prepared with AI assistance from the linked primary-source material. Claims and limitations should be checked against the original announcement or research.