OpenAI Signs $38 Billion Deal With Amazon's Cloud Computing Platform To Run Its AI Workloads

Hyderabad: The rapid advancement in AI technology has created unprecedented demand for computing power, leading AI firms to turn to cloud computing platforms to handle their workloads. Following the trend, ChatGPT-maker OpenAI has announced a multi-year, strategic partnership with Amazon's cloud computing platform, Amazon Web Services (AWS), to run and scale its core artificial intelligence (AI) workloads.

Starting immediately, the $38 billion deal will enable OpenAI to access AWS compute, comprising hundreds of thousands of NVIDIA GPUs, with the ability to expand to tens of millions of CPUs to rapidly scale agentic workloads.

OpenAI co-founder and CEO Sam Altman highlights that scaling frontier AI requires massive, reliable compute. He believes that their "partnership with AWS strengthens the broad compute ecosystem that will power this next era and bring advanced AI to everyone".

The agreement, which will have continued growth over the next seven years, will allow OpenAI to leverage AWS's experience running large-scale AI infrastructure at scale, with clusters topping 500K chips. While OpenAI starts utilising AWS compute, it will deploy full capacity before the end of 2026, expanding further into 2027 and beyond.

"AWS's leadership in cloud infrastructure, combined with OpenAI's pioneering advancements in generative AI, will help millions of users continue to get value from ChatGPT," OpenAI said in a blog post.

AWS and OpenAI announce multi-year strategic partnership (Image Credits: OpenAI)

From 2019 to 2023, OpenAI relied on Microsoft to provide computing power as part of a partnership between the two companies. The deal forced OpenAI to buy compute only from Microsoft, which was also their primary investor, unless the software major gave approval for computing deals with other companies. OpenAI's $39 billion compute deal with AWS follows the company's negotiations with Microsoft to let it purchase services from any cloud computing company without the tech giant's approval.

AWS infrastructure for OpenAI workloads

OpenAI says that AWS is building infrastructure deployment for OpenAI, featuring a sophisticated architectural design optimised for maximum AI processing efficiency and performance. It is clustering the NVIDIA GPUs, including both GB200s and GB300s, via Amazon EC2 UltraServers on the same network, enabling low-latency performance across interconnected systems. This allows OpenAI to efficiently run workloads with optimal performance.

It further explained that the GPU clusters have been designed to not only support various workloads, such as serving inference for ChatGPT and training next-generation models, but can also adapt to OpenAI's evolving needs.

“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 breadth and immediate availability of optimised compute demonstrates why AWS is uniquely positioned to support OpenAI's vast AI workloads."

Earlier this year, OpenAI made its open weight foundation models available on Amazon Bedrock, bringing them to millions of AWS customers for agentic workflows, coding, scientific analysis, mathematical problem-solving, and more. With its $38 billion deal with Amazon, OpenAI joins the likes of Anthropic, Stability AI, Salesforce, and Adobe that use AWS for their AI model workloads.