Elon Musk Announces Terafab, a $20 Billion Semiconductor Plant for Tesla and SpaceX
Vertical integration into chipmaking reflects growing anxiety about AI compute supply chains.
A Joint Venture Targeting One Terawatt of Annual Compute
On March 23, Elon Musk announced Terafab, a joint semiconductor fabrication facility developed by Tesla, SpaceX, and xAI. The planned plant, to be located in Austin, Texas, near Tesla's headquarters, will aim to produce custom chips optimized for electric vehicles, Optimus humanoid robots, and high-performance AI computing.
Musk described the project as necessary because global chip production levels are not sufficient to meet his companies' needs. He stated that current fabrication facilities produce only about two percent of what Tesla and SpaceX will eventually require.
Two Chip Types: Edge and Space
Terafab will produce two different types of semiconductors. One line will focus on edge inference workloads suitable for vehicles and robots. The second, more powerful line will be hardened for space environments to support SpaceX's planned network of orbital data centers.
The facility is expected to cost between 20 and 25 billion dollars and will be fully integrated, capable of making logic and memory chips as well as undertaking packaging and testing. Musk has no prior experience in semiconductor manufacturing, and the project carries significant execution risk.
Supply Chain Independence as Strategic Priority
The announcement reflects a broader industry anxiety about compute supply. Major cloud providers face capacity constraints, and GPU shortages continue to affect AI development timelines. Vertical integration into chipmaking represents an attempt to secure control over a critical bottleneck.
Whether Terafab reaches its stated production goals remains uncertain. The semiconductor industry is capital-intensive and operationally complex, and even established manufacturers face yield and timeline challenges.