Jane Edwards 
- AWS and NVIDIA have expanded their AI partnership to build federal AI factories
- The companies will bring Vera CPU infrastructure to AWS
- The 2026 FedCiv Summit will discuss AI, IT modernization and more
Amazon Web Services and NVIDIA have expanded their AI partnership to build AI factories for the U.S. government as part of a broader collaboration that will add 2 million additional graphics processing units to AWS’ global infrastructure by 2028.
Federal agencies continue to expand their use of cloud-based AI infrastructure to support mission and national security priorities. The Potomac Officers Club will host the 2026 FedCiv Summit on Oct. 29 to examine AI adoption, federal shared services, the OneGov strategy, digital labor, IT modernization and more. Reserve your spot now to gain firsthand insights from leaders shaping federal civilian modernization priorities.
What Does the Expanded Partnership Mean for Government Agencies?
In a joint release published Wednesday, AWS and NVIDIA said they will build AI factories for the U.S. government, including 100,000 GPUs deployed on secure AWS infrastructure to run federal and national security workloads. The collaboration is designed to support government agencies operating at Impact Level 6 and above, enabling agencies to deploy AI at scale for classified and mission-critical missions.
The government-focused buildout is part of a wider infrastructure expansion in which AWS will deploy 2 million additional NVIDIA Blackwell Ultra, Rubin and Rubin Ultra GPUs across its global infrastructure in 2027 and 2028, supporting agentic AI, enterprise automation, scientific discovery and physical AI workloads.
What Other Capabilities Are Included in the Expanded Collaboration?
Beyond the federal AI factories, AWS and NVIDIA are deepening integration across several areas of the AI stack. The companies plan to bring NVIDIA Vera CPU-based infrastructure to AWS, giving customers an additional compute option for agentic AI workloads that require high-performance CPUs alongside accelerated hardware.
AWS and NVIDIA will also extend NVIDIA NVLink Fusion with custom NVIDIA high-bandwidth memory, or NVHBM, to enhance performance and efficiency across Trainium chips and GPUs within a shared rack-scale architecture. The companies are integrating the NVIDIA platform with the AWS Nitro System and Elastic Fabric Adapter, or EFA, to strengthen security and reliability for GPU- and Trainium-based instances.
The collaboration further supports NVIDIA Nemotron open models on Amazon Bedrock and Amazon SageMaker, expanding model choice for customers, and accelerates data processing and vector indexing on Amazon EMR and Amazon OpenSearch using NVIDIA’s cuDF and cuVS CUDA-X libraries. Separately, Amazon Robotics is adopting NVIDIA’s physical AI platform to advance warehouse automation and next-generation robotics development.
What Did AWS & NVIDIA CEOs Say About the Expanded Collaboration?
AWS CEO Matt Garman said customers want flexibility to choose the AI tools that best fit their workloads while trusting that those tools work together seamlessly.
“That’s why we’ve invested deeply with NVIDIA to make AWS the best place to run NVIDIA AI technologies, optimizing performance across our infrastructure from networking and security to deployment. This expanded collaboration gives frontier labs, enterprises and governments even more ways to build and deploy AI on AWS,” Garman added.
Jensen Huang, founder and CEO of NVIDIA, said demand for the companies’ joint AI infrastructure has outpaced expectations since the partnership began 16 years ago.
“Now, we are expanding our partnership across the full stack — GPUs, CPUs, networking, open models and software — to make agentic and physical AI real at an unprecedented pace and scale that only AWS and NVIDIA can deliver. This expansion reflects customers’ demand for NVIDIA’s platform on AWS,” Huang stated.
How Does the Expanded Partnership Build on Federal AI Momentum of AWS & NVIDIA?
The expanded collaboration follows a string of moves by both companies to deepen their presence across the federal AI landscape.
AWS has been steadily expanding its government cloud footprint, including a potential $2.6 billion Department of Homeland Security Cumulus contract for commercial cloud services and a $50 billion investment to boost federal AI and supercomputing infrastructure across its Top Secret, Secret and GovCloud offerings. The company has also moved to streamline agency access to its technology through a $1 billion OneGov agreement with the General Services Administration aimed at accelerating IT modernization and AI adoption.
NVIDIA, for its part, has advanced its own government-focused push with its AI Factory for Government reference design built on its Blackwell architecture, and recently strengthened its public sector leadership with the appointment of Christopher George as vice president of public sector.
