Elon Musk has ruled out an operational role for TSMC in the chipmaking at Tesla's planned "Terafab," according to a report carried by the Waco Tribune-Herald.

The statement closes off one possible structure for Tesla's push into semiconductor manufacturing. Under the scenario Musk excluded, TSMC — the world's largest contract chipmaker and the sole producer of Nvidia's flagship AI accelerators — would have run day-to-day wafer operations at a Tesla-built facility rather than Tesla operating the plant itself.

What did Musk actually rule out?

Musk's position, as reported, is narrow but significant: TSMC will not hold an operational role in Terafab chipmaking. That leaves open several questions the report does not answer:

• Whether Tesla will build and run the Terafab entirely on its own
• Whether TSMC could still license process technology or provide design support without operating the fab
• Where the Terafab would be sited, at what capital cost, and on what timeline
• Which process nodes and wafer sizes Tesla would target for its own AI training silicon

The Terafab concept fits Musk's previously stated ambition to scale up compute for Tesla's AI program, including training of its autonomous-driving and humanoid-robot models. Building fab capacity in-house, rather than buying exclusively from merchant foundries, would give Tesla direct control over a critical supply constraint.

Why the TSMC question mattered

Any involvement by TSMC in a Tesla fab would have carried weight well beyond one customer relationship. TSMC runs the most advanced logic processes in the industry and its capacity is heavily booked by AI chip demand. A Terafab operated with TSMC's know-how would effectively have extended the foundry's reach into a US-built, Tesla-controlled site.

Musk ruling that out signals Tesla intends to keep operational control of its own chipmaking if and when the Terafab proceeds. It also sharpens the competitive picture: Tesla's in-house AI chip efforts, including its Dojo training compute program, have so far existed alongside — not instead of — purchases of Nvidia GPUs.

How does this fit the wider fab race?

Leading-edge fabrication remains extraordinarily capital-intensive, with new advanced fabs typically costing tens of billions of dollars and years to bring online. Tesla has not confirmed a capacity target, node, or start date for any Terafab, so the project should be read as a stated direction rather than a committed build with defined output.

The context is a broader scramble for advanced compute. AI developers across the industry are confronting long lead times for top-end accelerators, and several big spenders have explored ways to secure dedicated capacity. Musk's decision to exclude TSMC from operations suggests he prefers owned control over a foundry partnership model, even though in-house leading-edge manufacturing carries substantial execution risk.

For now, no capacity figures, investment amounts, or process-roadmap details for the Terafab have been disclosed, and Musk's statement does not commit to a construction date. The next concrete signal to watch is whether Tesla announces a site, capital budget, or equipment orders for the Terafab — the steps that would convert the concept into a scheduled fab project.