Source
AWS and NVIDIA expand partnership for next-gen AI infrastructure
AWS and NVIDIA are expanding their strategic partnership to deploy 2 million additional GPUs by 2027–2028 and integrate new CPU and networking technologies. The collaboration aims to scale AI infrastructure for enterprises, frontier labs, and the U.S. government.
Why it matters
This expansion gives businesses and government agencies more computing power and hardware choices to develop complex AI applications like agentic AI and scientific discovery. It allows users to optimize their AI workloads by choosing the most efficient chip and memory configurations.
The details
Amazon's Annapurna Labs is integrating NVIDIA's NVLink Fusion and custom high-bandwidth memory into Trainium chips. NVIDIA Nemotron open models are currently available via Amazon Bedrock and SageMaker. Additionally, Amazon Robotics is utilizing NVIDIA's Jetson, Omniverse, and Isaac platforms to advance warehouse automation.
What's next
AWS plans to deploy the additional 2 million NVIDIA GPUs across its global infrastructure throughout 2027–2028.
Get the weekly recap
The stories like this one, picked and explained — once a week, straight to your inbox.