Speaking in Austin, Musk described the Terafab concept as an effort to materially expand computing supply for artificial intelligence, robotics and space applications, areas he indicated may require far more silicon than is currently available. Estimates referenced by analysts suggest the project could involve $5 trillion to $13 trillion in capital spending, supporting hundreds of fabrication plants and targeting up to one terawatt of annual compute capacity, a scale that may rival or exceed tod
[Illustrative image] ... Washington: ... by the world's richest person. The "Terafab", a manufacturing facility based near Austin, Texas, will aim to produce one terawatt of computing power per year, Musk said....
Positive Sentiment: TeraFab / AI chip narrative draws bullish analyst support — Wedbush and others continue to call Tesla an emerging AI/semiconductor play after the Terafab chip-factory effort, which boosts the company’s long-term “physical AI” thesis. Dan Ives: Terafab Upside
Elon Musk unveiled an ambitious plan Saturday night to build what he calls the world’s largest semiconductor manufacturing complex in Austin, Texas — a joint venture between Tesla, SpaceX and xAI he is calling “Terafab.”
Terafab is expected to accelerate Tesla and SpaceX's AI development by ensuring chip and memory supply, helping the companies overcome current constraints and enabling ambitious AI projects like FSD, Cybercab, and Optimus.
ASML, Applied Materials, Lam Research, and KLA dominate critical manufacturing process steps. Additionally, Micron, Samsung, and SK Hynix provide advanced memory dies. Therefore, TeraFab must place multi-billion-dollar orders years in advance. So far, public filings show no firm purchase orders for EUV or immersion lithography tools.
Musk admitted that partners like Samsung and TSMC couldn't meet the scaling speed requirements. Therefore, Tesla decided to build its own TeraFab to ensure a reliable chip supply. Experts warn the project faces major challenges, particularly the supply of EUV lithography machines from ASML ...
A 2nm fab requires multiple High-NA EUV lithography systems from ASML. These machines cost hundreds of millions of dollars and have waitlists that span years. We have not seen any confirmed orders from Tesla in the ASML backlog yet. The clock for a fab does not start when a press release is issued; it starts when the equipment slots are secured. Furthermore, a facility like Terafab requires thousands of specialized engineers.
A recently issued research note ... by 2027. Ives argues that the groundwork is already actively being laid and points to the newly announced Terafab project as the ultimate catalyst for this historic combination....
Positive Sentiment: Wedbush analyst Daniel Ives reiterates a bullish view on Tesla and calls the Terafab chip push “an emerging AI powerhouse,” supporting upside thesis for Tesla’s AI/robotics roadmap and the stock’s re-rating.
Musk concluded his Terafab presentation with a video showing future data center satellites being built on the Moon and launched using the electromagnetic mass driver.
Terafab is a $25 billion chip manufacturing facility that Musk said will be capable of producing 1 terawatt per year in computing power. At that rate of output, it would be the largest semiconductor fab in the world, by a lot.
Construction is seen north of the Tesla Giga Texas factory in Austin on Wednesday, March 25, 2026. CEO Elon Musk said Tesla and SpaceX are planning the world’s largest chip manufacturing plant, called Terafab, but said this site would be used ...
Tesla has announced plans to build a major semiconductor fabrication facility called Terafab in Austin, Texas, in partnership with SpaceX, to secure advanced AI chip supplies for its vehicles, robotics, and data center projects.
Speaking in Austin, Musk described the Terafab concept as an effort to materially expand computing supply for artificial intelligence, robotics and space applications, areas he indicated may require far more silicon than is currently available. Estimates referenced by analysts suggest the project could involve $5 trillion to $13 trillion in capital spending, supporting hundreds of fabrication plants and targeting up to one terawatt of annual compute capacity, a scale that may rival or exceed tod