Terafab aims for mass chip production by late 2028, supplier delivery speeds up

The wait for new chips from the Terafab project may be shorter than previously thought. Elon Musk announced yesterday that major equipment suppliers are moving faster to ship machinery for the joint Tesla and SpaceX venture. This acceleration supports a goal to begin volume production of chips by the third or fourth quarter of 2028. The facility is located in Gibbons Creek, Texas. Musk described this schedule as highly aspirational, noting it comes with significant uncertainty.

During a live broadcast of SpaceX's earnings call on X, Musk expressed gratitude to partners for the speed. He stated that the dates are very aggressive. The company plans to manufacture its first experimental test chip at a research facility in Austin, Texas, around the second quarter of 2027. This step serves as a precursor to full-scale operations at the main plant.

Terafab is designed to be a vertically integrated manufacturing hub. It will produce logic chips, memory, and advanced packaging in a single location. The project aims to generate one terawatt of AI computing power annually. Musk claimed that once fully operational, the commercial value of this output could rival the GDP of a top-20 global country. For comparison, Switzerland, currently ranked 20th, has a GDP of approximately $1.14 trillion.

The facility relies on extreme ultraviolet lithography systems. It will use Intel's 14A manufacturing node technology. Intel joined the venture in April 2026 to help refactor silicon fabrication. ASML, the sole commercial supplier of these critical EUV tools, is a key partner. Unofficial reports suggested the first EUV tool would arrive in mid-2027, but it is unclear if that aligns with the new expedited timeline Musk mentioned.

Beyond ASML and Intel, the project involves numerous other suppliers. American firms Applied Materials, Lam Research, and KLA Corporation are providing front-end processing and control machinery. Japan's Tokyo Electron is also part of the supply chain. Additionally, Tesla and SpaceX are in talks with about ten South Korean companies. These include Hanmi Semiconductor for advanced packaging and HPSP for annealing equipment, as well as providers of specialized piping and valves.

Musk outlined broader computing goals during the same broadcast. He expects his AI companies to have 2.5 to 3 gigawatts of active capacity by the end of this year. He is confident about adding 10 more gigawatts in 2027. This would bring the total to 13 gigawatts by the end of next year. To reach this milestone, SpaceX and xAI are procuring next-generation NVIDIA architecture. A cluster of 440,000 Nvidia Vera Rubin GPUs is planned to be online by the end of the second quarter of 2027.

Critics remain skeptical of the one-terawatt ambition. Research firm Gartner estimates total global AI-optimized data center capacity at roughly 44 gigawatts as of late 2026. This represents only about 4.4% of Terafab's planned annual output. While global AI capacity has grown rapidly since 2022, the scale of Terafab's target is immense. The project aims to reduce dependence on external foundries like TSMC and supply chips for xAI clusters, self-driving cars, Optimus robots, and SpaceX's orbital data centers.