OpenAI plans to produce and deploy its own AI chips next year with support from semiconductor company Broadcom, focusing the hardware on internal operations rather than selling to outside customers. The news was first reported by the Financial Times and relayed by TipRanks. According to TipRanks, the move follows months of rumors about OpenAI’s custom chip efforts and aims to reduce reliance on external suppliers.
OpenAI’s strategy and rationale
Reports earlier this year suggested OpenAI was exploring new suppliers to help develop its first generation of in-house AI processors as it looked to lessen dependence on Nvidia. OpenAI is following in the footsteps of companies such as Alphabet, Meta, and Amazon, which have also developed custom chips to meet rising demand from large-scale AI workloads.
Training and operating large language models requires significant computing power. By producing its own chips, OpenAI aims to lower costs, reduce dependence on external providers, and improve long-term scalability, TipRanks reported.
Broadcom’s “fourth customer” and future AI revenue
After reporting strong third-quarter earnings, Broadcom CEO Hock Tan said the company has secured a large “fourth customer” for its AI infrastructure products. Tan added that AI-related revenue growth is expected to “improve significantly” in FY26, supported by more than $10 billion in infrastructure orders from this new customer. While Tan did not disclose the name, Financial Times sources cited by TipRanks suggest the client is OpenAI. Broadcom already benefits from three undisclosed large clients that have been major contributors to its AI revenue growth.
Analyst context from TipRanks
TipRanks noted prior reporting that OpenAI’s chip push aligns with broader industry trends among major tech firms building custom silicon. The outlet emphasized that OpenAI’s planned chips are intended for internal use, supporting the company’s own AI workloads rather than external sales.
According to TipRanks, the Financial Times report underscores ongoing momentum in AI infrastructure investment and the strategic value of in-house hardware for scaling large language model operations.
