OpenAI, the company behind ChatGPT, is taking a major step toward AI hardware independence by developing its first in-house artificial intelligence chip in partnership with US semiconductor giant Broadcom. According to a report by the Financial Times, the new silicon is expected to debut in 2026 and will initially be used for OpenAI’s own workloads rather than sold externally.
Reducing Reliance on Nvidia GPUs
The move signals OpenAI’s strategy to reduce dependence on Nvidia, whose GPUs dominate the AI chip market and power most large language models (LLMs). Reports also suggest OpenAI has engaged with TSMC for chip fabrication, while continuing to leverage processors from AMD and Nvidia to manage growing demand for computing resources.
Broadcom’s $10 Billion AI Infrastructure Orders
During an earnings call, Broadcom CEO Hock Tan revealed the company secured more than $10 billion in AI infrastructure orders from a new client, widely believed to be OpenAI. Tan emphasized that AI revenue growth will “improve significantly” in fiscal 2026 as these partnerships mature, though he did not directly name OpenAI.
Earlier this year, Tan hinted that Broadcom was collaborating with four potential new customers on custom AI chips, in addition to its existing top clients.
By designing its own silicon, OpenAI joins the ranks of Google (TPU), Amazon (Inferentia and Trainium), and Meta, all of whom have developed proprietary AI hardware to cut costs and boost performance. OpenAI’s first-generation chip design is reportedly nearing completion and will soon be sent to TSMC for production.
Competing in a Resource-Intensive AI Market
As the race to build more powerful AI models intensifies, the need for efficient and scalable computing has become critical. OpenAI’s push into custom AI hardware not only reduces risks associated with reliance on a single supplier but also positions the company to better scale future AI models while keeping costs under control.
With AI demand soaring worldwide, OpenAI’s Broadcom-powered chip marks a pivotal step in reshaping the semiconductor landscape and the future of AI innovation.
