AWS announces new AI Factories to reduce infrastructure barriers for public, private sector

LAS VEGAS — Amazon Web Services and NVIDIA are combining forces to create specialized artificial intelligence “factories” for their customers looking to access advanced computational resources without building their own expensive infrastructure. 

Announced on Tuesday at the AWS re:Invent conference in Las Vegas, the AWS AI Factories look to reduce the infrastructure burden for public and private sector customers that want to run advanced AI calculations, but lack the necessary components of the AI tech stack, like data storage or supercomputing chips. 

“As governments and large organizations seek to scale AI projects, some are turning to the concept of an ‘AI factory’ to address their unique sovereignty and compliance needs,” the press release announcing the partnership says. “AWS AI Factories combine the latest AI accelerators including cutting-edge NVIDIA AI computing and Trainium chips, AWS high-speed, low-latency networking, high-performance storage and databases, security, and energy-efficient infrastructure together with comprehensive AI services like Amazon Bedrock and Amazon SageMaker so customers can rapidly develop and deploy AI applications at scale.” 

AWS and NVIDIA software and hardware will work in tandem in the AI Factories to supply customers with the AI infrastructure components they lack, while working with existing customer environments. The goal is to reduce the capital investment such entities need to fully scale an AI-ready digital environment. 

“By combining NVIDIA’s latest Grace Blackwell and Vera Rubin architectures with AWS’s secure, high-performance infrastructure and AI software stack, AWS AI Factories allow organizations to stand up powerful AI capabilities in a fraction of the time and focus entirely on innovation instead of integration,” said Ian Buck, vice president and general manager of Hyperscale and HPC at NVIDIA, in the press release.

To further incentivize government customers to take advantage of AI Factories, AWS said that the spaces are designed to meet Unclassified, Sensitive, Secret, and Top Secret security clearance levels. 

In addition to AWS and NVIDIA, Saudi-based AI company HUMAIN is also included in the partnership. AWS will be building a novel “AI Zone” in Saudi Arabia, which will include up to 150,000 AI chips, including GB300 GPUs, and select AWS infrastructure –– all within a HUMAIN data center. 

The partnership takes advantage of new authorizations for international chip sales. The GB300 is prohibited from being sold to Chinese-affiliated entities, but the Commerce Department approved their sales to the United Arab Emirates last month, and specifically HUMAIN, contingent on participating companies “meeting rigorous security and reporting requirements.”

“The AI factory AWS is building in our new AI Zone represents the beginning of a multi-gigawatt journey for HUMAIN and AWS,” Tareq Amid, CEO of HUMAIN, said in a statement. “From inception, this infrastructure has been engineered to serve both the accelerating local and global demand for AI compute.”