Terafab Targets 1 Terawatt Of Annual AI Compute Output By 2028

A joint manufacturing project by Tesla and SpaceX is aiming to produce high volumes of advanced chips by late 2028, a target that the companies' CEO admits carries "big error bars." Elon Musk said suppliers of critical semiconductor equipment, including ASML, are speeding up equipment deliveries to support this timeline. The schedule would see the Gibbons Creek, Texas facility begin volume production in the third or fourth quarter of 2028. "These are very aggressive dates, and we're very appreciative of ASML and all of our other suppliers in accelerating equipment deliveries in order to achieve this outcome," Musk said.

The discussion took place during a live broadcast of SpaceX's Q3 2026 earnings call on October 20. As part of the roadmap, SpaceX plans to assemble its first experimental test chip at a research fabrication facility in Austin, Texas, around the second quarter of 2027. The project, known as Terafab, is designed to manufacture one terawatt of AI computing power per year from a single building. It is intended to fabricate logic, memory, and advanced packaging in one location. Once operational, the facility's commercial value is projected to rank within the top 20 global country GDPs. For context, Switzerland currently holds the 20th position with a GDP of approximately $1.14 trillion.

Musk also outlined broader plans for AI computing capacity during the broadcast, stating that his companies expect to have 2.5 to 3 gigawatts of computing capacity active by the end of 2026. He expressed high confidence regarding an additional 10 gigawatts coming online in 2027, noting he "would be surprised" if the total does not reach 13 gigawatts by the end of that year. To meet this objective, SpaceX and xAI are acquiring NVIDIA's next-generation Vera Rubin architecture. The companies announced plans to bring a 440,000-GPU training cluster online by the end of the second quarter of 2027.

By comparison, other large-scale initiatives are following different schedules. Project Stargate, which has a planned capacity of 10 gigawatts and has been restructured, is expected to reach full capacity in the 2030s. Meta's 5 gigawatt project, Hyperion, follows a similar timeline. Regarding the fabrication technology required for Terafab, the facility will utilize extreme ultraviolet lithography systems based on Intel's 14A manufacturing node. Intel joined the venture in April 2026 to help co-develop and refactor silicon fabrication technology. ASML is the sole commercial supplier of the $400 million EUV lithography systems needed for high-volume production.

Other equipment providers involved in the project include Applied Materials, Lam Research, KLA Corporation, and Tokyo Electron. These companies will supply front-end processing, metrology, and process control machinery. Tesla and SpaceX are also in discussions with approximately 10 South Korean equipment makers, including Hanmi Semiconductor and HPSP, as well as firms that provide specialized piping components and process valves. Regarding the delivery schedule, reports indicated that the first EUV tool was slated for mid-2027, though it remains unclear if this is the expedited timeline referred to during the broadcast. The facility aims to mitigate future shortfalls of processors from external foundries and will supply chips for xAI clusters, vehicles, humanoid robots, and orbital data centers.

Skeptics have questioned the validity of the one terawatt ambition. Research firm Gartner estimates the total global AI-optimized data center capacity is roughly 44 gigawatts as of late 2026, representing 4.4% of the facility's planned output. According to EpochAI, global AI computing capacity has grown by roughly 3.3 to 3.4 times per year since 2022. The timeline remains highly aspirational, and the delivery of the first EUV machine is a critical factor for the schedule.

The facility will fabricate logic, memory, and advanced packaging internal architectures in one place. The project's massive planned output has mandated various collaborations across the semiconductor supply chain. Tesla and SpaceX are also in talks with about 10 South Korean equipment makers, including Hanmi Semiconductor, known for advanced packaging equipment, and HPSP, a high-pressure hydrogen annealing equipment manufacturer, as well as firms providing specialized piping components and process valves.