The company plans to collaborate with Taiwan’s TSMC for manufacturing, with initial production tests expected to take several months. The chip will be fabricated using TSMC’s advanced 3-nanometer process, paving the way for mass production in 2026.

A team of 40 engineers, led by former Google chip expert Richard Ho, is spearheading the project. The group is working closely with Broadcom to develop a chip capable of both training and running AI models, though the initial rollout is expected to be limited.

The relatively small team has raised some questions. Industry experts note that full-scale chip development programs typically require hundreds of engineers, making OpenAI’s initiative appear modest compared to similar projects by Google or Amazon.

Sources familiar with the project suggest that OpenAI primarily views the chip as leverage in negotiations with other suppliers. While Nvidia currently dominates the AI chip market, tech giants like Amazon, Microsoft, and Meta* have been striving to develop their own hardware for years.

This push comes amid surging demand for AI chips. Amazon, Meta*, and Microsoft plan to invest tens of billions in AI infrastructure, while OpenAI itself is involved in the $500 billion Stargate infrastructure initiative.