- What changed
- Amazon Web Services and NVIDIA released a guide for building physical artificial intelligence model factories using NVIDIA Cosmos 3 on Amazon SageMaker HyperPod clusters, focusing on synthetic data and evaluation.
- Why you should care
- Engineering teams building physical artificial intelligence systems require continuous pipelines rather than isolated training jobs, and this integration helps maintain high graphics processing unit utilization during complex synthetic data generation and evaluation workflows.
- Your move
- Watch. Teams building physical artificial intelligence applications can use this reference architecture to set up persistent training and evaluation pipelines on Amazon Web Services infrastructure.
- What to watch next
- Watch for further verified reporting or independent evidence.
- Event
- release
- Event date
- Sep 4, 2026
- Relevant to
- Cloud infrastructure engineers, Machine learning operations professionals, Robotics developers
