Elon Musk announced Sunday that SpaceX and Tesla plan to construct two advanced semiconductor manufacturing facilities in Austin, Texas, aimed at supporting next-generation technologies across his companies. The project, known as “Terafab,” would consist of two separate fabrication plants, each dedicated to producing a single specialized chip design.

Speaking at a presentation in Austin and in follow-up remarks on X, Musk said one facility would manufacture chips for Tesla’s electric vehicles and its Optimus humanoid robots, while the second would focus on producing high-performance processors intended for artificial intelligence systems, including those deployed in space-based data centers.

The announcement builds on Musk’s earlier statement that Tesla would eventually need to produce its own AI chips to meet internal demand. The involvement of SpaceX, however, had not previously been disclosed. The aerospace company is also preparing for a potential public listing that could value it at approximately $1.75 trillion, following its recent merger with Musk’s artificial intelligence venture, xAI.

Musk emphasized that existing global semiconductor production is insufficient to meet the anticipated needs of his expanding portfolio of companies. He said current chip output would account for only a fraction of future demand, warning that without projects like Terafab, supply constraints could limit growth.

While expressing appreciation for current suppliers, including Samsung, Taiwan Semiconductor Manufacturing Company, and Micron, Musk indicated that reliance on third-party manufacturers would not be sustainable long term given the scale of computing power required.

According to Musk, the Terafab complex is expected to eventually produce up to one terawatt of computing capacity annually, a figure that would exceed the roughly half-terawatt currently generated across the United States. He did not provide a construction timeline but acknowledged the ambitious scope of the project.

Musk also highlighted the technical challenges involved in developing chips for space-based applications, noting that processors designed for satellites must withstand harsher environmental conditions, including higher operating temperatures and radiation exposure.