The plan is to start with a smaller, highly advanced fabrication facility capable of producing and testing a wide range of chips, before expanding to a larger operation.

Musk argues the semiconductor industry isn't scaling fast enough to meet his companies' growing demand for AI and robotics, so he sees building his own supply as necessary. His long-term goal is to support massive computing capacity—eventually reaching a terawatt annually—though he hasn't provided a timeline.

Yahoo writes that the project would likely sit near Tesla's Austin headquarters and could produce cutting-edge chips, potentially at the 2-nanometer level. One set of chips would power vehicles, robotaxis, and humanoid robots, while another, more powerful line would be designed for space-based computing used by SpaceX and xAI.

Despite widespread concern about chip shortages, it's unusual for companies to build their own fabs due to the enormous cost and complexity. Musk acknowledged existing suppliers can't fully meet Tesla's future needs as it shifts toward AI-driven products.

He also outlined broader ambitions, including space-based data centers powered by satellite networks. A prototype "mini" satellite could deliver about 100 kilowatts, with future versions reaching megawatt levels. These efforts are tied to SpaceX's planned IPO and larger vision of expanding computing infrastructure beyond Earth.

Overall, the Terafab project reflects Musk's push to vertically integrate chip production while supporting his longer-term goals in AI, robotics, and space technology.