OpenAI and Broadcom Collaborate on Custom AI Chip Development
OpenAI has entered into a significant partnership with Broadcom to produce its first in-house AI processors, aiming to secure the necessary computing power as demand for its services rises.
The companies announced on Monday that OpenAI will design the chips, while Broadcom will manage development and deployment, with production slated to begin in the latter half of 2026.
They plan to introduce 10 gigawatts of custom AI chips, equivalent to the electricity required by over 8 million U.S. households, and about five times the power output of the Hoover Dam. All new systems will operate on Broadcom’s Ethernet and other networking equipment.
This move poses a direct challenge to Nvidia’s InfiniBand technology, which currently dominates AI data centers. Following the announcement, Broadcom’s stock increased by more than 10%.
OpenAI’s collaboration with Broadcom is part of a strategy to establish a stronger infrastructure base. Historically, OpenAI has relied heavily on Nvidia’s powerful chips for training and operating its models.
However, the surge in demand for these chips has led to supply bottlenecks and increased costs. OpenAI’s self-designed chips aim to provide control over performance, availability, and expenses.
Analysts suggest that this deal will not immediately threaten Nvidia’s dominant position in the AI accelerator market, as designing, scaling, and manufacturing chips from scratch is complex and costly.
Despite this, the demand for custom processors has made Broadcom a major beneficiary of the AI boom, with its stock rising nearly six-fold since the end of 2022.
OpenAI’s ambitious chip development plan is expected to be completed by the end of 2029. According to Nvidia CEO Jensen Huang, constructing a single one-gigawatt data center can cost between $50 billion and $60 billion.
Financing such a large project will likely involve multiple strategies, including funding rounds, pre-orders, strategic investments, credit facilities, and support from Microsoft.
The partnership with Broadcom follows a series of large chip investments by OpenAI, including a recent 6-gigawatt AI chip supply deal with AMD.
This collaboration positions OpenAI among a growing group of technology leaders working to develop custom chips, potentially reducing its reliance on Nvidia and strengthening its strategic position in the global AI market.