By Priya Raman

TSMC is evaluating an expansion of its Texas manufacturing footprint, CEO C.C. Wei said, but he attached one firm condition: any so-called Terafab facility must be operated by TSMC itself. The statement defines the boundary of the company's US ambitions. It also signals what TSMC will not concede.

The term Terafab refers to a hypothetical very large-scale integrated fabrication complex, a plant sized to serve an entire customer base rather than a single product line. Wei's phrasing matters because it separates two things that US policymakers often merge: where a fab sits, and who runs it. TSMC has now stated its position on the second question. A site can move to Texas. Operational control cannot.

What did Wei actually commit to?

Very little, on close reading. He confirmed that TSMC is evaluating Texas expansion options. Evaluation is not construction, and construction is not commitment. TSMC's existing US site in Arizona took years from announcement to volume production, a useful yardstick for how slowly evaluation converts to output in American semiconductor manufacturing.

The company has not disclosed:

• A site selection or land acquisition in Texas
• A capital expenditure figure for the expansion
• A construction timeline or production start date
• The process nodes (the chip generations) a Texas plant would run

Each absence narrows what the announcement means in practice. Investors should treat this as an option, not a plan.

Why does the Terafab operator clause matter?

Wei's insistence on TSMC operation addresses a recurring tension in US industrial policy. Washington wants advanced logic capacity (the chips that power processors) on American soil, funded heavily by the CHIPS and Structures incentives framework. But TSMC's competitive advantage lives in its process discipline and yield management (the share of manufactured chips that pass quality tests). Handing operation of a Terafab to a partner or a joint venture would export the building while retaining the knowledge that makes the building valuable.

Wei said the Terafab must be operated by TSMC. That single sentence tells partners, customers and government negotiators where the negotiation stops. Any Texas expansion will carry TSMC management, TSMC process engineers and TSMC accountability for output.

Who wants this, and who pays?

The report does not specify which customers or government bodies prompted the Texas evaluation. It does not name a subsidy amount. It does not say whether the evaluation covers one fab or several. The silence is itself information: at this stage, TSMC benefits from signaling flexibility in the US without locking in cost.

Texas competes on land availability, power costs and a lighter regulatory burden than some coastal states. Arizona, by contrast, offered proximity to an established TSMC workforce. A Texas site would start that workforce accumulation from a lower base, and the announcement does not address labor supply at all.

What happens next?

Watch for concrete markers: a land purchase, a board-approved capex line, or a named incentive package from Texas authorities. Until one appears, the Terafab condition is a negotiating position, not a factory. Wei has told the market what TSMC might build and under whose flag. He has not yet told anyone when.