TSMC is considering a multibillion-dollar semiconductor campus in Texas, according to a report by The Business Times, as the world's largest contract chipmaker weighs how to add manufacturing capacity for AI-related silicon on US soil.

The reported plan centers on a dedicated campus rather than a single fab building, which signals a scale of investment that would extend the company's existing American manufacturing position. TSMC has not confirmed the investment amount, the site selection, or a construction timeline, and the company routinely evaluates capacity options before committing capital.

The Texas proposal arrives as demand for AI accelerators and high-performance computing devices continues to absorb a growing share of leading-edge wafer capacity. Customers designing AI chips depend on advanced process nodes, and any new campus would presumably be evaluated against that demand profile. The Business Times report links the potential expansion directly to the production of more AI chips.

What a campus-level investment implies

A campus configuration typically involves multiple fabrication facilities, support buildings, and utility infrastructure built out in phases. That structure allows a manufacturer to add capacity incrementally as orders firm up, rather than committing the entire capital outlay at once. For TSMC, a Texas campus would represent a further step in a geographic expansion strategy that has already taken the company beyond its Taiwan manufacturing base.

The multibillion-dollar figure attached to the reported deliberations is consistent with the capital intensity of leading-edge semiconductor manufacturing. Advanced fabs routinely require investments in the tens of billions of dollars when tooling, cleanroom construction, and utility systems are included.

Texas has an established semiconductor manufacturing presence, and the state has previously attracted major chipmaking investments. Site decisions of this magnitude typically hinge on utility availability, water supply, workforce pipelines, and the terms of government incentives.

The AI demand context

AI chips are among the most capacity-hungry products in the semiconductor portfolio. They use large die sizes, advanced packaging, and high-bandwidth memory integration, all of which consume more wafer and assembly capacity per unit of revenue than many legacy products. A manufacturer expecting sustained AI-driven demand needs to plan capacity years ahead, because a new fab takes several years from groundbreaking to volume production.

For TSMC specifically, the decision on a US AI-focused campus would follow the company's stated approach of locating capacity close to major customers where the economics support it. The company has committed to US production capacity in Arizona, and the reported Texas deliberations suggest the company sees additional American demand that its announced capacity cannot fully serve.

Open questions

Several material details remain unconfirmed. The report does not specify which process technologies the Texas campus would host, when construction could begin, or how the investment would be phased. It also does not state whether the project would depend on US government incentive terms or on commitments from specific AI chip customers.

TSMC's capital planning follows customer demand signals, and a project of this scale would ordinarily require long-term purchase commitments before final approval. Until the company formally announces the investment, the Texas campus remains an option under evaluation rather than a committed project.

Die Signal will continue to track the story as TSMC's capacity plans for AI chips in the United States develop.