OpenAI moves deeper into chip development
OpenAI is preparing to release its first in-house artificial intelligence chip in 2026. According to a Financial Times report, the company will partner with U.S. semiconductor giant Broadcom to bring the project to life. The move highlights OpenAI’s growing presence in the AI ecosystem and its strategy to control more of its technology stack.
Focus on internal use, not customers
The report noted that OpenAI plans to use the chip internally instead of selling it to customers. This aligns with the company’s heavy reliance on computing power to train and run large language models, including ChatGPT.
Shaping a path beyond Nvidia
For years, OpenAI has relied heavily on Nvidia’s GPUs. However, rising demand and high costs pushed the company to explore alternatives. In February, Reuters reported that OpenAI was finalizing the design of its first in-house chip and planned to send it for fabrication to Taiwan Semiconductor Manufacturing Company (TSMC).
OpenAI has also been working with Advanced Micro Devices (AMD) chips alongside Nvidia to diversify its supply chain. Broadcom’s involvement marks a key milestone in this journey.
Broadcom’s AI bet grows bigger
Broadcom CEO Hock Tan recently revealed that the company secured more than $10 billion in AI infrastructure orders from a new customer, without naming it. Analysts believe OpenAI could be that client, given the reported collaboration.
Tan also said Broadcom expects AI revenue growth to improve significantly by fiscal 2026. Earlier this year, the company hinted at multiple new clients developing custom chips with its help, beyond its existing major partners.
A familiar trend among tech giants
OpenAI’s step mirrors strategies of other big tech firms. Google, Amazon, and Meta have already built their own chips to power AI workloads. Developing in-house silicon allows companies to optimize performance, reduce costs, and secure supply amid growing demand.
OpenAI’s expanding footprint
OpenAI has already claimed more than 100 million users for its Copilot AI tools. The new chip project signals the company’s intent to further strengthen its infrastructure. By moving closer to custom silicon, OpenAI joins the league of global players shaping the next phase of AI development.
