OpenAI is investing $10 billion in a partnership with Broadcom to develop new AI chips. These chips are set to launch in 2026, reported the Financial Times and Wall Street Journal.
The goal is to reduce OpenAI’s reliance on Nvidia, the main supplier of AI processors for cloud services like Amazon Web Services, Google, Microsoft, and Oracle.
OpenAI set to start mass production of its own AI chips with Broadcom https://t.co/eM3SceFVsI
— Financial Times (@FT) September 5, 2025The chips will be used inside OpenAI for training and running AI models like ChatGPT. Broadcom, based in California, hinted at a major new contract recently. This deal is a big boost for the company.
Nvidia’s processors have powered much of the AI growth. For example, Oracle plans to buy $40 billion worth of Nvidia chips for its Stargate project. Still, OpenAI joins other big firms like Google, Amazon, and Microsoft in making its own chips to diversify supply chains.
OpenAI links up with Broadcom to produce its own AI chips https://t.co/eWQJhLdMTX
— Ars Technica (@arstechnica) September 5, 2025NVIDIA’s sales rose 56% last quarter. With the possible easing of trade restrictions under the Trump administration, Nvidia’s market could grow even more.
Read: Oracle to Invest $40 Billion in Nvidia Chips for OpenAI’s Stargate Data Center
OpenAI’s work with Broadcom signals a key change in the AI field. It challenges Nvidia’s lead and could drive new chip innovations. The 2026 launch may reshape AI development.
