Twelve wafer fabs, an advanced packaging line, and an R&D center already sit in Arizona. That is the capacity baseline any second U.S. park must clear before a single cleanroom column is poured. The constraint is not demand; it is whether a second state can absorb the power, water, and skilled-labor load that twelve fabs already impose on one.

The digitimes report, relayed through a single X post on September 30, 2026, uses the phrase "second U.S. manufacturing park" for the Texas project. It implies a cluster: multiple wafer lines, packaging, possibly R&D, mirroring the Arizona structure. The document does not say how many fabs, at what node, or with what wafer-start capacity.

What the document does fix is the Arizona reference: 12 wafer fabs, advanced packaging, an R&D center. A production node is not named for either site. An investment figure does not appear. A power requirement is not stated. The September 30 timestamp and 908 views on the X post are metadata, not engineering data.

Die Brief's read: this is a planning rumor, not a procurement trigger. No TSMC statement is cited, and the report does not establish that ground will be broken in Texas. What it does establish is that the Arizona cluster has reached a scale where a second U.S. site is being discussed as a structural option rather than a contingency. For a buyer or a fab planner, nothing changes until a node and a capacity number appear.

A Texas location is not named. Wafer capacity, power draw, and official TSMC confirmation are all absent. The digitimes article, as quoted in the X post, attributes the report to the fab-services supply chain, which is industry parlance for the construction and utilities vendors who see early site surveys. That is a signal, not a commitment.

Watch for a TSMC investor-relations filing or a state-level economic-development announcement. Until one of those appears, the Texas project is a constraint on TSMC's internal capacity planning, not on the supply chain. The 12 Arizona fabs remain the only confirmed U.S. wafer capacity, and they are the number that matters to anyone ordering substrates or packaging capacity in the next two quarters.

The document does not prove TSMC will build in Texas; it proves the Arizona cluster is large enough to make a second U.S. site a structural question. Until a node and wafer-start number appear, this is internal capacity planning, not a supply-chain event.

Single-source, unconfirmed. The 12-fab Arizona baseline is the only hard number in the document. No node, no capacity, no power draw, no TSMC statement. The fab-services-chain attribution suggests early site surveys but not a commitment. Downstream planning is unaffected.