2 min read
OpenAI has entered a long-term partnership with Amazon Web Services (AWS) to scale its next-generation AI models using AWS infrastructure. The seven-year deal, valued at $38 billion, is effective immediately and will enable OpenAI to expand its compute capacity significantly.
Under the agreement, OpenAI will utilize Amazon EC2 UltraServers equipped with hundreds of thousands of Nvidia GPUs. These systems can scale up to tens of millions of CPUs, providing the power required for advanced generative AI workloads. The deployment will be completed before the end of 2026, with an option to expand further from 2027 onward.
According to the official statement, the partnership will help OpenAI “rapidly expand compute capacity while benefiting from the price, performance, scale, and security of AWS.” The company added that AWS’s experience running large-scale AI infrastructure and managing clusters exceeding 500,000 chips was a key factor in the decision.
The architecture integrates Nvidia’s GB200 and GB300 GPUs on a unified network to deliver low-latency performance across interconnected systems. This setup will allow OpenAI to train and run its models more efficiently, ensuring faster and more reliable performance for millions of ChatGPT users.
