1 Terawatt: The Massive Gamble Behind Terafab's Chip Ambitions

The central tension in Elon Musk’s latest semiconductor push lies in the gap between a single facility’s production target and the current scale of the global AI industry. On October 20, 2026, during a live broadcast of SpaceX’s Q3 earnings call on X, Musk announced that critical suppliers like ASML are expediting shipments for Terafab, a joint venture between Tesla and SpaceX. The project aims to manufacture one terawatt (TW) of AI computing power per year from a single location in Texas. While Musk describes the timeline as "highly aspirational," the sheer magnitude of the goal creates a stark contrast with existing infrastructure standards.

The numbers underpinning this ambition are difficult to contextualize. Gartner estimates that the total global AI-optimized data center capacity as of late 2026 is roughly 44 gigawatts (GW). This means Terafab’s planned annual output of 1 TW would be equivalent to approximately 22.7 times the entire current global capacity. To achieve this, the facility in Gibbons Creek, Texas, targets volume chip production by the third or fourth quarter of 2028. Musk acknowledged the aggressive nature of these dates, noting they carry "big error bars" and that the company is "very appreciative" of suppliers accelerating deliveries to make this possible.

The project involves a complex web of partnerships across the semiconductor supply chain. Intel joined the venture in April 2026 to co-develop silicon fabrication technology, specifically providing access to its 14A manufacturing node. ASML, the sole commercial supplier of the $400 million extreme ultraviolet (EUV) lithography systems required for leading-edge production, is considered the most critical partner. Unofficial sources suggest the first EUV tool was slated for delivery in mid-2027, though it remains unclear if this aligns with the expedited timeline Musk referenced. Other key suppliers include American firms Applied Materials, Lam Research, and KLA Corporation, as well as Japan’s Tokyo Electron.

Beyond lithography, the supply chain extends to front-end processing and specialized components. Tesla and SpaceX are in negotiations with approximately 10 South Korean equipment makers, including Hanmi Semiconductor for advanced packaging and HPSP for high-pressure hydrogen annealing. The facility is designed to be vertically integrated, fabricating logic, memory, and advanced packaging in one place. This internal architecture aims to mitigate future shortfalls of cutting-edge AI processors from external foundries like TSMC, ensuring a dedicated supply for Musk’s xAI clusters, self-driving vehicles, Optimus robots, and SpaceX’s orbital data centers.

The stakes are defined by both commercial value and technological feasibility. Musk stated that once fully operational, the commercial value of Terafab’s chip output would place the facility within the top 20 global country GDPs. For perspective, the 20th-ranked economy is Switzerland, with a GDP of around $1.14 trillion. This valuation assumes the facility successfully achieves its multi-billion-dollar, vertically integrated goal without significant delays. The project also supports broader AI compute goals, with Musk expecting 2.5 to 3 GW of capacity active by the end of 2026 and a total of 13 GW by the end of 2027.

To support these compute targets, SpaceX and xAI are procuring next-generation NVIDIA hardware, including a plan to bring a 440,000-GPU "Vera Rubin" training cluster online by the end of Q2 2027. However, experts have questioned the validity of the 1 TW Terafab ambition. While global AI computing capacity has grown by 3.3 to 3.4 times per year since 2022, according to EpochAI, scaling a single facility to such an extreme level presents unique logistical challenges. The success of Terafab will depend on whether the accelerated equipment deliveries can overcome the inherent tightness in the global semiconductor supply chain.