AWS and NVIDIA Expand Partnership for Next-Gen AI Infrastructure
AWS and NVIDIA have announced an expanded partnership to deliver next-generation AI infrastructure. As part of this agreement, AWS will deploy an additional 2 million NVIDIA GPUs across its global infrastructure in 2027-2028.
This move complements Amazon's own custom silicon, allowing customers to choose the best compute for their specific workloads. The companies are working together on various fronts, including AI factories, CPUs, networking, open models, data processing, and robotics.
NVIDIA Vera CPUs will be coming to AWS, providing an additional compute option for agentic AI workloads. This will enable customers to configure AI infrastructure in a way that suits their needs.
The companies are also building AI factories for the U.S. government, delivering NVIDIA's AI stack on AWS's secure infrastructure for federal and national-security workloads classified at Impact Level 6 (IL6) and above.