TSMC staff express doubts over the company's decision to build a $40B factory in Arizona, arguing the project distracts from R&D, showing TSMC's tricky position
Employee doubts are rising about Taiwan Semiconductor Manufacturing Company's $40 billion investment in an Arizona factory.
Context & Ripple Effects
The Arizona project began modestly as a $12B 5nm fab announced in 2020 under White House pressure to onshore chip production, then ballooned into a $40B two-fab commitment within TSMC's broader $200B US investment plan. Execution has been the running story since: the plant fell behind schedule badly enough that several hundred more workers were flown in from Taiwan, while recruiters struggled because the company's rigid culture turned off local candidates for the planned 4,500 hires.
First-order effects
- TSMC engineers are openly questioning whether management is diverting capital and attention from the R&D pipeline that keeps it ahead of Samsung and Intel, an internal-friction problem layered on top of an already-delayed construction timeline.
Second-order effects
- With TSMC forecasting years of margin dilution as overseas fabs ramp, its reported plan to raise advanced and mature chip prices by up to 10% in 2027 shifts the cost of politically driven capacity onto customers like Apple and Nvidia.
Third-order effects
- If the pattern holds, leading-edge foundries will permanently run two books — process leadership at home and subsidized capacity abroad — forcing Washington to soften the kind of subsidy conditions TSMC called 'unacceptable' or watch foreign champions deprioritize US fabs.
The trend: Semiconductor manufacturing is splitting into a geopolitical layer of government-subsidized overseas fabs and an R&D core, with foundry margins and talent becoming the contested ground between them.