Intel will remain involved in Terafab, the semiconductor production venture backed by SpaceX, Tesla, and xAI, according to comments from Intel chief executive Lip-Bu Tan. Tan spoke to Bloomberg before an Intel event in Japan marking the company's presence there. The confirmation follows an April announcement that Terafab would use Intel's next-generation 14A fabrication process to make chips. While the full terms of the collaboration remain undisclosed, the general understanding is that Intel will license its manufacturing technology to Terafab, though it is unclear whether Intel will also help ramp production on the node.
Elon Musk simultaneously pushed back on rumors that his companies would invest in Terafab and then partner with TSMC to build and operate the fab. In a post on X, Musk said his companies will build and run the fab themselves and that there should be zero doubt about it. He suggested TSMC could sublease part of Terafab if it wanted, but said nothing more than that would happen. The proposal is an unusual arrangement for TSMC, the world's largest foundry, which owns and operates dozens of advanced 300-mm fabs globally.
TSMC has its own production facilities and aggressive plans for roughly ten new fab phases in the coming years, so it has little need to sublease capacity, a building, or a site from a third party. Semiconductor fabs are also not generic facilities. If Terafab is initially equipped and qualified around Intel's 14A process, converting part of the facility for TSMC's A14, A13, or A12 processes planned for the back half of the decade would require different process flows, equipment configurations, materials, automation, qualification, and contamination control. TSMC could theoretically install a largely independent production line in the same enormous building, but that would resemble TSMC operating its own fab inside a Musk-controlled site rather than renting spare capacity.
IP and contractual complications add further questions. An Intel 14A license to Terafab would almost certainly include extensive provisions governing access to Intel intellectual property, process information, equipment configurations, and personnel. Placing TSMC engineers and manufacturing operations in the same facility could create contractual and IP-isolation issues even if TSMC never directly touches the Intel production line. IP protection would be extraordinary at Terafab, complicating its operation.
Operational liability is another concern. If two chipmakers end up in the same fab building or campus, responsibility for downtime becomes messy. Should a Terafab-owned power distribution system, water plant, chemical system, or cooling installation fail and take production offline, determining operational responsibility and financial liability for potentially hundreds of millions of dollars of affected wafers would be considerably more complicated than at a single-company site. If Terafab equips enough cleanroom space by the end of the decade to satisfy Tesla, SpaceX, and xAI while leaving substantial capacity for TSMC, that would raise questions about whether the project overbuilt capacity in its initial phases.
Musk's statements on X are deliberately vague, making them hard to interpret. A conceptual problem runs through the proposal. If TSMC installs its own tools, hires its own employees, and brings process technology and production management while Terafab provides land, shell, and infrastructure, calling it a sublease undersells what would be a TSMC fab embedded in the Terafab campus. If instead Terafab provides tools and infrastructure and TSMC merely operates them, that looks closer to the operational involvement Musk appears to deny. Intel has confirmed its involvement, and Musk wants TSMC involved as well, but the latest comments raise more questions than they answer.
More company and startup news from TechManNews.
