Terafab: Why Musk Chose Intel Over TSMC for a $25 Billion Chip Megafactory
The Terafab project represents a fundamental shift in how America's most prominent technology entrepreneur sources silicon. Rather than relying on TSMC — the default choice for virtually every advanced chip design in the world — Musk's $25 billion Austin facility will use Intel as its fabrication backbone. The decision is not merely commercial; it reflects the growing strategic premium on domestic manufacturing amid escalating US-China trade tensions and tariff uncertainty. Intel's 18A process node, now in high-volume production with its novel RibbonFET and PowerVia architecture, gave Musk a credible American alternative at the leading edge.
Terafab's ambitions are staggering in scope: 100,000 wafer starts per month and 1 terawatt per year of compute output. The facility will produce energy-efficient edge-inference processors for Tesla's Full Self-Driving system, Cybercab fleet, and Optimus robots, alongside radiation-hardened chips for SpaceX and the Starshield military satellite program. For Intel, the deal provides something money alone cannot buy — a guaranteed, high-volume foundry customer that validates the entire IFS (Intel Foundry Services) business model. As Intel stated: "Our ability to design, fabricate, and package ultra-high-performance chips at scale will help accelerate Terafab's aim to produce 1 TW/year of compute to power AI and robotics." The partnership transforms Intel from a company trying to sell foundry services into one with a marquee anchor tenant.



