Intel will remain part of Elon Musk's Terafab semiconductor venture, Intel CEO Lip-Bu Tan told reporters on Wednesday. The confirmation came ahead of an Intel event in Japan and follows the April announcement that Terafab would use Intel's next-generation 14A fabrication process. The exact structure of that collaboration is still unclear, though the general expectation is that Intel will license its manufacturing technology to Terafab.
Around the same time, Musk pushed back on rumors that his companies would invest in Terafab only for TSMC to build and operate the fab. "No, we will build and run the fab," he wrote on X. "Maybe TSMC subleases part of the Terafab if they want, but nothing more than that." Tom's Hardware points out that this is an unusual proposal: TSMC already has its own advanced fabs and aggressive expansion plans, so it has little reason to sublease capacity from a third party.
The outlet also raises several practical concerns. A fab built around Intel's 14A process would not be easily converted for TSMC's A14, A13, or A12 flows, which require different equipment, materials, automation, and contamination control. Sharing a campus could also create IP-isolation problems, complicate liability if shared infrastructure fails, and raise questions about whether Terafab is overbuilding capacity. In short, the latest comments reaffirm Intel's involvement but leave the shape of any TSMC arrangement far from settled.