OpenAI’s First Artificial Intelligence Chip Development with Broadcom

OpenAI is embarking on the production of its initial artificial intelligence chip in collaboration with the US semiconductor powerhouse Broadcom, as reported by the Financial Times, which cited sources knowledgeable about the situation. This new silicon is anticipated to be operational next year and will be used internally, not made available for external customers.

Reasons Behind the Move

This initiative underscores OpenAI’s efforts to lessen its reliance on Nvidia, whose GPUs currently dominate the industry for training and running extensive language models. Previously, Reuters indicated that OpenAI is collaborating with both Broadcom and Taiwan Semiconductor Manufacturing Co. (TSMC) in creating the chip, while also utilising processors from AMD alongside Nvidia to meet the rising demand for computational resources.

Project Confirmation

A source associated with the project has confirmed that the design of OpenAI’s first-generation chip is approaching completion and will soon be dispatched to TSMC for manufacturing.

Broadcom’s Outlook on AI Revenue

Broadcom’s CEO Hock Tan stated on Thursday that the company anticipates significant growth in AI revenue during fiscal 2026, having secured over $10 billion in infrastructure contracts from a new client, widely believed to be OpenAI. Tan informed investors during an earnings call that a new prospect placed a firm order last quarter, thus qualifying as a customer, without disclosing the identity of the buyer.

Collaboration with Potential Customers

Earlier this year, Tan suggested that Broadcom is closely engaged with four prospective clients on custom chip projects, in addition to its three established major clients.

Broader Industry Trends

OpenAI’s strategy reflects similar initiatives by technology giants like Google, Amazon, and Meta, all of whom have designed proprietary chips to manage costs and performance for their AI applications more effectively. By launching its own hardware, OpenAI is poised to achieve greater flexibility in scaling future models while ensuring tighter cost management.

Intensifying Competition in AI

This development comes at a time when competition within the AI sector is escalating, as developing increasingly sophisticated models demands considerable computing power. OpenAI has conveyed that its long-term approach involves diversifying supply chains to diminish dependence on a singular chip supplier.