OpenAI Partners with Broadcom to Launch First AI Chip in 2026
OpenAI is preparing to launch its first artificial intelligence chip in 2026 through a strategic partnership with U.S. semiconductor giant Broadcom, according to a Financial Times report citing sources familiar with the matter. The ChatGPT creator plans to use the custom chip exclusively for internal operations rather than making it available to external customers, marking a significant step in the company’s effort to reduce dependency on external chip suppliers.
Strategic Move to Reduce Nvidia Dependence
This development represents OpenAI’s latest attempt to diversify its chip supply chain and control costs as demand for AI computing power continues to surge. The company currently relies heavily on Nvidia chips to train and operate its generative AI systems, which require substantial computational resources to produce human-like responses to user queries.
The custom chip initiative aligns with OpenAI’s broader strategy that began taking shape last year when the company started working with Broadcom and Taiwan Semiconductor Manufacturing Co (TSMC) to develop in-house silicon solutions. The company has also been incorporating AMD chips alongside Nvidia hardware to meet growing infrastructure demands.
Industry-Wide Trend Toward Custom Silicon
OpenAI’s move mirrors similar strategies adopted by major technology companies seeking greater control over their AI infrastructure. Google, Amazon, and Meta have all developed custom chips to handle AI workloads as the industry grapples with unprecedented computational requirements for training and operating advanced AI models.
Broadcom CEO Hock Tan recently indicated strong momentum in the AI sector, announcing that the company expects “significant improvement” in artificial intelligence revenue growth for fiscal 2026. The semiconductor manufacturer has secured over $10 billion in AI infrastructure orders, including from a new major customer that Tan declined to identify during an earnings call.
This chip development represents a crucial milestone in OpenAI’s evolution from an AI research organization to a comprehensive technology company capable of controlling its entire computational stack.