AWS to provide AI infrastructure deployed in customer's own data centres

Amazon Web Services (AWS) has announced AWS AI Factories, a new offering that provides enterprises and governments with dedicated AWS AI infrastructure deployed in their own data centers.

AWS AI Factories combine the latest AI accelerators, including NVIDIA AI computing and Trainium chips, AWS high-speed, low-latency networking, high-performance storage and databases, security and energy-efficient infrastructure, together with AI services like Amazon Bedrock and SageMaker AI.

AWS AI Factories operate like a private AWS Region that gives secure, low-latency access to compute, storage, database, and AI services.

This approach lets customers leverage existing data center space and power capacity they've already acquired and gives access to AWS AI infrastructure and services - from the latest AI chips for training and inference to tools for building, training, and deploying AI models.

It also provides managed services that offer access to foundation models without having to negotiate separate contracts with model providers.

With the NVIDIA-AWS AI Factories integration, AWS customers have access to the NVIDIA accelerated computing platform, full-stack NVIDIA AI software, and thousands of GPU-accelerated applications.

The AWS Nitro System, Elastic Fabric Adapter (EFA) petabit-scale networking, and Amazon EC2 UltraClusters support the latest NVIDIA Grace Blackwell and the next-generation NVIDIA Vera Rubin platforms.

In the future, AWS will support NVIDIA NVLink Fusion high-speed chip interconnect technology in next-generation Trainium4 and Graviton chips, and in the Nitro System.

AWS AI Factories can run workloads across all government classification levels: Unclassified, Sensitive, Secret, and Top Secret.

"These AI factories operate exclusively for each customer, and it helps them with that separation, maintaining the security and reliability they get from AWS, while also meeting stringent compliance and software engineering requirements,"  said Matt Garman, CEO of AWS in his keynote speech at AWS re:Invent 2025 event in Las Vegas, Nevada.

"We're super excited to see what these AI factories unlock for customers."

Jason Pollock travelled to AWS re:Invent 2025 as a guest of AWS.