Image: Getty Images/ For illustrative purposes
OpenAI has signed a multi-year strategic partnership with Amazon Web Services (AWS) valued at $38bn to run and scale its advanced artificial intelligence workloads on AWS infrastructure.
Under the agreement, OpenAI will immediately begin using AWS compute capacity, including hundreds of thousands of NVIDIA GPUs connected through Amazon EC2 UltraServers, with the ability to scale to tens of millions of CPUs.
All capacity is targeted to be deployed by the end of 2026, with expansion expected into 2027 and beyond.
Highlights of the OpenAI-AWS tieup
The deal will give OpenAI access to AWS’s large-scale AI infrastructure, which includes clusters of more than 500,000 chips.
The setup will support both inference for ChatGPT and the training of next-generation models, with systems designed for low-latency performance and scalable efficiency.
“Scaling frontier AI requires massive, reliable compute,” said OpenAI co-founder and CEO Sam Altman. “Our partnership with AWS strengthens the broad compute ecosystem that will power this next era and bring advanced AI to everyone.”
“As OpenAI continues to push the boundaries of what’s possible, AWS’s best-in-class infrastructure will serve as a backbone for their AI ambitions,” AWS CEO Matt Garman said.
The partnership follows earlier collaborations between the two firms. OpenAI’s foundation models are already available on Amazon Bedrock, AWS’s AI service, where they are used by customers such as Bystreet, Comscore, Peloton, Thomson Reuters, Triomics, and Verana Health for applications including coding, data analysis, and scientific research.
AWS said the new infrastructure will use NVIDIA GB200 and GB300 chips to handle OpenAI’s growing compute requirements, reinforcing the company’s position as a preferred provider for large-scale AI workloads.
Read: AWS’ Tanuja Randery on supporting EMEA’s digital transformation and AI ambition
