OpenAI is advancing towards finalizing its first in-house artificial intelligence (AI) chip design, aiming to reduce its reliance on Nvidia’s hardware. The company plans to complete the design in the coming months and send it to Taiwan Semiconductor Manufacturing Co. (TSMC) for fabrication using advanced 3-nanometer technology. Mass production is targeted for 2026.
The initial deployment of these custom chips will be on a limited scale, primarily to run AI models. The design is expected to feature high-bandwidth memory and extensive networking capabilities. This move is seen as a strategic effort to strengthen OpenAI’s negotiating position with other chip suppliers and to test the feasibility of expanding its chip development efforts.
OpenAI’s chip design team, led by former Google TPU engineer Richard Ho, has grown to 40 members. The company is collaborating with Broadcom in this endeavor. This initiative aligns with a broader trend among tech giants like Microsoft and Meta, who are investing heavily in AI infrastructure and seeking alternative chip solutions due to rising costs and reliance on Nvidia
OpenAI is getting closer to launching its in-house chip
OpenAI is advancing towards launching its in-house AI chip, aiming to finalize the design within the next few months and commence mass production by 2026. The company plans to send the finalized design to Taiwan Semiconductor Manufacturing Co. (TSMC) for fabrication using advanced 3-nanometer technology. The chip is expected to feature a systolic array architecture, high-bandwidth memory, and extensive networking capabilities, similar to Nvidia’s AI chips.
This initiative is part of OpenAI’s strategy to reduce its reliance on Nvidia GPUs, which currently dominate the AI hardware market. The in-house chip development is led by a team of about 40 engineers, including former Google TPU engineer Richard Ho, and is being developed in collaboration with Broadcom. Initially, the chips will be used on a limited scale primarily to run AI models.
OpenAI’s move reflects a broader trend among tech giants like Microsoft and Meta, who are investing heavily in AI infrastructure and developing custom chips to optimize performance and manage costs.