SpaceX is reportedly preparing to spend at least $55 billion on a Texas factory that would push the company into AI chip manufacturing at enormous scale.

Reports from The New York Times and CNBC, citing details in a public hearing notice filed in Grimes County, say the project centers on a plant known as “Terafab” in the Austin area. The filing suggests the price tag will be steep even by the standards of a company already known for betting big on hardware, infrastructure, and long timelines.

A reported $55 billion chip plant would mark a dramatic expansion of SpaceX beyond rockets and satellites and into the core machinery powering the AI boom.

The move fits a broader scramble across the tech industry to secure the chips that train and run advanced AI systems. Demand for that hardware keeps climbing, and companies now see control over supply as a strategic advantage, not just a manufacturing detail. If SpaceX follows through, it would join a crowded and capital-intensive fight where scale, speed, and access to power matter as much as design.

Key Facts

  • Reports indicate SpaceX plans to invest at least $55 billion in an AI chip plant.
  • The project is reportedly called “Terafab.”
  • A public hearing notice filed in Grimes County points to the planned facility.
  • The plant would be located in the Austin, Texas, area, according to reports.

What remains unclear matters just as much as the headline figure. The public reporting does not yet answer when construction would begin, what kind of chips the plant would make, or how the output would connect to Elon Musk’s wider AI ambitions. Those gaps leave room for caution, but they do not shrink the significance of the plan itself: a project this large would signal long-term confidence in domestic chip production and in AI demand that shows no sign of cooling.

The next steps will likely come through local hearings, regulatory filings, and more detailed disclosures about the site and its purpose. For Texas, the project could deepen the state’s role as a center for advanced manufacturing. For the AI industry, it would sharpen a trend already underway: companies no longer want to just buy the future’s most important hardware — they want to build it themselves.