OpenAI, the creator of ChatGPT, announced on Monday a strategic partnership with semiconductor giant Broadcom to co-design and manufacture custom processors for artificial intelligence.
The collaboration, set to launch next year, aims to deliver 10 gigawatts of computing power, roughly equivalent to the energy needed to sustain a major city. The move marks another major step in OpenAI’s push to secure its dominance in the generative AI revolution that began with ChatGPT’s debut in November 2022.
Led by CEO Sam Altman, OpenAI has recently inked several major deals with global tech leaders, including Nvidia, AMD, Oracle, Samsung, and SK Hynix. These partnerships underscore the company’s effort to expand its infrastructure and meet the surging global demand for AI services.
Despite booming investor enthusiasm, analysts warn that the massive infrastructure build-up, including energy-intensive data centers and high-powered AI chips, could strain electricity providers and inflate financial risks reminiscent of the late 1990s dot-com bubble.
While financial details of the Broadcom deal remain undisclosed, Broadcom shares jumped nearly 10 percent following the announcement, mirroring gains seen by AMD and Oracle after their own partnerships with OpenAI.
By developing its own chips, OpenAI hopes to optimize processing efficiency and performance beyond what off-the-shelf processors can deliver. “Partnering with Broadcom is a critical step in building the infrastructure needed to unlock AI’s potential and deliver real benefits for people and businesses,” said Altman.
Broadcom described the partnership as “a pivotal moment” in AI’s evolution, as the custom processors are expected to power OpenAI’s expanding network of data centers worldwide.
What you should know
OpenAI’s partnership with Broadcom highlights a growing trend among tech giants to design proprietary AI hardware, reducing dependence on traditional chipmakers and boosting efficiency. The deal also underscores the escalating race to secure the energy and infrastructure required to sustain global AI demand.
