Mon Oct 13 13:50:00 UTC 2025: Okay, here’s a news article summarizing the provided Reuters text:
**OpenAI Partners with Broadcom to Develop Custom AI Chips in Massive Power Deal**
**San Francisco, CA -** OpenAI, the creator of ChatGPT, has forged a partnership with Broadcom to develop its first in-house artificial intelligence processors, signaling a major push to secure the computing power needed to fuel its rapidly growing AI services.
The deal, announced Monday, will see OpenAI designing the chips, with Broadcom handling development and deployment, starting in the second half of 2026. The initial rollout will involve a staggering 10 gigawatts of custom chips, a level of power consumption comparable to the needs of over 8 million U.S. households or five times the power output of the Hoover Dam.
This agreement is one of many of the massive investments in AI chips highlighting the exponential growth of the AI industry and the insatiable demand for computing power. Broadcom shares surged over 7% in early trading following the news.
The partnership places OpenAI alongside tech giants like Google and Amazon, which are also developing custom chips to meet AI demand and reduce reliance on Nvidia.
While financial details were not disclosed, OpenAI CEO Sam Altman stated the Broadcom partnership is a “critical step in building the infrastructure needed to unlock AI’s potential.”
The Broadcom partnership comes on the heels of OpenAI announcing a 6-gigawatt AI chip supply deal with AMD, days after disclosing that Nvidia plans to invest up to $100 billion in the startup.
Broadcom’s role in custom AI chips has made it one of the biggest winners of the generative AI boom, with its stock price rising nearly six-fold since the end of 2022.
The deployment of the new custom chips is expected to be completed by the end of 2029, utilizing Broadcom’s Ethernet and networking gear, giving the company an edge over smaller rivals and challenging Nvidia’s InfiniBand networking solution.
First Piper 