OpenAI is preparing to produce its first artificial intelligence chip next year in collaboration with U.S. semiconductor giant Broadcom (AVGO.O), the Financial Times reportedA, citing people familiar with the matter.
According to the report, the chip will be used internally to power OpenAI’s systems rather than being offered to external customers. Reuters could not independently verify the report, while both OpenAI and Broadcom declined to comment outside business hours.
OpenAI, best known as the creator of ChatGPT, requires vast computing resources to train and run its AI models. Last year, Reuters reported that the company was working with Broadcom and Taiwan Semiconductor Manufacturing Co (TSMC) to design its first in-house chip. At the same time, OpenAI has been using AMD and Nvidia processors to cope with growing infrastructure demands, while exploring ways to diversify supply and reduce costs.
In February, Reuters said OpenAI was finalizing the design of its in-house silicon, with plans to send it to TSMC for fabrication. The move is aimed at reducing reliance on Nvidia, which dominates the AI chip market.
Broadcom CEO Hock Tan told analysts on Thursday that AI revenue growth for fiscal 2026 is expected to “improve significantly,” after the company secured more than $10 billion in AI infrastructure orders from a new unnamed customer. The firm placed a confirmed order last quarter, officially qualifying as a Broadcom client.
Tan has previously indicated that four potential customers, in addition to three existing major clients, were working with Broadcom on custom chip development.
OpenAI’s push into chipmaking mirrors similar strategies by Google, Amazon, and Meta, all of which have designed their own custom silicon to meet the surging demand for AI computing power.
