Amazon Web Services and Nvidia announced an expanded partnership to deploy more than 2 million additional Nvidia GPUs across AWS's global infrastructure by 2028. The collaboration spans multiple Nvidia product lines, including Blackwell Ultra, Rubin, and Rubin Ultra GPUs, as part of a broader effort to enhance AI compute capacity.
The agreement also includes the delivery of 100,000 GPUs on AWS's secure infrastructure for federal and national-security workloads, supporting classified applications at "Impact Level 6 and above." AWS will be the first major cloud provider to offer compute instances with Nvidia RTX PRO 4500 Blackwell Server Edition GPUs, integrated into Amazon EC2 G7 instances.
Beyond hardware, the companies are collaborating on AI infrastructure development, including Nvidia's Vera CPU, designed for high-performance AI workloads. Nvidia's Nemotron family of open models will be made available on Amazon Bedrock and Amazon SageMaker, while GPU-accelerated data processing is set to expand on Amazon EMR using Amazon EC2 G7 instances.
The partnership extends into robotics, with Amazon Robotics leveraging Nvidia's physical AI platforms—Jetson, Omniverse libraries, and Isaac—for next-generation robotic systems. AWS CEO Matt Garman emphasized the goal of providing customers with seamless integration of AI tools, stating, "Customers want the freedom to choose the best tools for their AI workloads, and they want confidence that everything works seamlessly together."
Nvidia Founder and CEO Jensen Huang highlighted the scale of demand, noting, "NVIDIA and AWS have built one of the great growth engines of the AI era, and demand is running ahead of every forecast."
