OpenAI, a Longtime Nvidia Customer, Plans to Buy Google’s AI Chips

OpenAI, a longtime customer of Nvidia, plans to buy artificial intelligence chips from Google, according to a report by The Information. The move would mark a notable expansion of OpenAI’s hardware sourcing beyond Nvidia, whose graphics processing units have become central to training and running many advanced AI systems.

Google develops its own AI accelerators, commonly known as Tensor Processing Units, for use in data centers and machine-learning workloads. Such chips are designed to handle the intensive computing tasks required by large AI models, including training and inference.

The reported plan highlights the growing importance of access to specialized computing hardware as AI companies expand their products and services. Nvidia remains a major supplier of AI chips, while Google is among the large technology companies that have developed custom silicon for internal cloud and AI operations.

The Information’s report did not, in the material provided, specify the scale, timing, commercial terms, or exact chip models involved in OpenAI’s planned purchases. Any arrangement would place Google’s AI hardware more directly in OpenAI’s broader computing supply chain.