TSMC's total U.S. investment commitment — already announced at $265 billion — may grow further, according to a GuruFocus report, extending what is already the largest single foreign semiconductor manufacturing commitment in American history.

The $265 billion figure, confirmed by TSMC earlier this year, covers the company's fab cluster in Phoenix, Arizona, where the world's largest contract chipmaker is building out multiple leading-edge facilities. The original plan called for six fabs on the site; the question raised by the report is whether additional phases, wafer capacity or support facilities will push total spending above the announced envelope.

TSMC's Arizona operations anchor the U.S. attempt to rebuild leading-edge logic manufacturing on domestic soil. The first fab, Fab 21 phase 1, entered volume production on the N4 process — TSMC's 4-nanometer-class node — supplying customers including Apple and Nvidia. Subsequent phases are slated to bring N3 (3-nanometer-class) and eventually N2 (2-nanometer-class) production, plus advanced packaging capacity, to U.S. soil for the first time at scale.

The scale matters commercially. A $265 billion commitment across six fabs dwarfs Intel's announced U.S. buildout plans and Samsung's Taylor, Texas project, and it anchors TSMC's answer to customer and political pressure to diversify production away from Taiwan. Apple, Nvidia, AMD and Qualcomm have all committed to sourcing chips from Arizona, giving the site a customer base that mirrors TSMC's advanced-node order book in Hsinchu.

Whether the program grows beyond $265 billion depends on variables the report does not fully resolve: customer demand for U.S.-made leading-edge wafers, the pace of AI-accelerator orders, the terms of remaining CHIPS Act and tariff-related incentives, and construction and equipment cost inflation in the U.S. versus Taiwan. Historical precedent cuts both ways. TSMC raised its Arizona commitment from an initial $12 billion single-fab plan announced in 2020, to $40 billion for two fabs, to $65 billion for three, before the March announcement of $100 billion in additional investment brought the total to $165 billion — and further revisions lifted it to the current $265 billion. Each revision followed customer commitments, not speculative capacity planning.

The company has consistently framed the expansion as customer-driven. Arizona capacity is booked largely by U.S. customers who want geographic diversification of advanced-node supply, and expansion decisions have tracked demand signals from those accounts rather than TSMC's own capacity targets.

Geopolitics shapes the commercial logic. Washington's tariff regime and export controls have raised the effective cost of fabbing leading-edge chips exclusively in Taiwan, while CHIPS Act grants — TSMC received up to $6.6 billion in direct funding for its first three Arizona fabs — subsidize a portion of U.S. construction costs. For TSMC, the calculus is whether U.S.-made wafers priced above Taiwanese equivalents remain competitive for customers whose boards demand supply-chain resilience.

Cost remains the friction point. U.S. fab construction and operating costs run well above Taiwan's, and TSMC has acknowledged the premium. Customers have so far absorbed it, betting that diversified leading-edge supply reduces exposure to Taiwan Strait risk more than the added cost harms margins. Any further expansion of the $265 billion program would signal that this bet is deepening — more capacity, more nodes, and a larger share of TSMC's leading-edge output stamped "made in America."

No revised figure, timeline or additional fab count has been confirmed by TSMC. Should the company formalize an increase, expect the announcement to follow the established pattern: tied to named customer commitments, phased over years, and concentrated in Phoenix. Competitive dynamics in leading-edge foundry would then shift further, as Intel's foundry push and Samsung's Taylor ramp compete for the same finite pool of U.S.-based advanced-node demand.