OpenAI signs $38 billion compute deal with Amazon, partnering with cloud leader for first time

OpenAI has entered into a monumental agreement with Amazon Web Services (AWS), securing $38 billion in cloud computing capacity. This marks OpenAI's first collaboration with the leading cloud infrastructure provider and signifies a shift away from its previous reliance on Microsoft. The partnership, announced on Monday, allows OpenAI to utilize AWS infrastructure immediately, leveraging hundreds of thousands of Nvidia graphics processing units (GPUs) in the United States, with plans for further capacity expansion in the years to come. Following the announcement, Amazon's stock saw a notable increase of approximately 5%. The initial phase of this agreement will focus on existing AWS data centers, with Amazon committing to develop additional infrastructure tailored to OpenAI's needs. Dave Brown, vice president of compute and machine learning services at AWS, remarked, "It's completely separate capacity that we're putting down. Some of that capacity is already available, and OpenAI is making use of that." OpenAI has been actively pursuing partnerships recently, having announced around $1.4 trillion in agreements with major firms such as Nvidia, Broadcom, Oracle, and Google. This surge in partnerships has raised questions from skeptics about the potential for an AI bubble and the feasibility of meeting ambitious industry promises. Previously, OpenAI held an exclusive cloud partnership with Microsoft, which had invested a total of $13 billion since backing the company in 2019. However, in January, Microsoft shifted to a non-exclusive arrangement, allowing OpenAI to explore other cloud partnerships. OpenAI has also secured agreements with Oracle and Google, but AWS stands out as the dominant player in the cloud market. "Scaling frontier AI requires massive, reliable compute," stated OpenAI CEO Sam Altman in the announcement. "Our partnership with AWS strengthens the broad compute ecosystem that will power this next era and bring advanced AI to everyone." Despite this new deal, OpenAI remains committed to Microsoft, recently announcing plans to purchase an additional $250 billion in Azure services. For Amazon, this partnership is not only significant due to its size but also because the cloud giant has invested heavily in OpenAI's rival, Anthropic. Amazon is in the process of constructing an $11 billion data center campus in Indiana specifically for Anthropic's workloads. AWS CEO Matt Garman emphasized, "The breadth and immediate availability of optimized compute demonstrates why AWS is uniquely positioned to support OpenAI's vast AI workloads." In the latest earnings report, Amazon reported over 20% year-over-year revenue growth at AWS, surpassing analyst expectations, although Microsoft and Google reported even faster growth rates of 40% and 34%, respectively. The current agreement with OpenAI specifically involves Nvidia chips, including two popular Blackwell models. There is potential for integrating additional chip technology in the future. While Amazon's custom-built Trainium chip is currently utilized by Anthropic, details regarding its use with OpenAI remain undisclosed. This new infrastructure will support both inference tasks, such as providing real-time responses for ChatGPT, and the training of next-generation AI models. OpenAI will have the flexibility to expand its AWS resources over the next seven years, although plans beyond 2026 are yet to be determined. OpenAI's foundational models are already accessible through Bedrock, AWS's managed service for AI systems, and various companies are leveraging OpenAI models for tasks ranging from coding to scientific analysis. This arrangement solidifies OpenAI's status as a customer of AWS, with a clear commitment to purchasing compute capacity. As the most valued private AI company, this agreement positions OpenAI for future growth, signaling its operational maturity and independence. Altman has previously indicated that an initial public offering (IPO) is likely on the horizon, a sentiment echoed by CFO Sarah Friar, who views the recent corporate restructuring as a crucial step towards that goal.