OpenAI is making a bold move to reduce its dependence on Nvidia by developing its own custom AI chip. This ambitious project, led by former Google chip designer Richard Ho, involves a team of 40 engineers working with Broadcom on the processor design. The chip will be manufactured by TSMC, the same company that produces Nvidia's chips, using its 3-nanometer process technology.
OpenAI's first chip will primarily focus on running AI models, also known as "inference," rather than training them. The company plans to deploy these chips across its operations, with mass production potentially starting in 2026.
This move is part of a larger trend, with major tech companies investing heavily in AI infrastructure. Microsoft plans to invest $80 billion in 2025, while Meta has set aside $60 billion for the next year. OpenAI itself has announced a $500 billion "Stargate" infrastructure project to build new AI data centers in the US.
