TSMC Struggles to Meet AI Chip Demand Despite $165 Billion U.S. Expansion
Updated
Updated · The Verge · Jun 4
TSMC Struggles to Meet AI Chip Demand Despite $165 Billion U.S. Expansion
3 articles · Updated · The Verge · Jun 4
Summary
C.C. Wei said TSMC still cannot fully meet surging U.S. customer demand for AI semiconductors, warning the company can only support so much even as it tries to avoid becoming a bottleneck.
AI adoption is driving the squeeze across chips and memory, with Deloitte projecting the semiconductor market could reach $1 trillion by 2027.
U.S.-based output will not quickly close the gap: Wei said it could take a "very long time" to satisfy customers through American production.
TSMC has already opened one Arizona fab and plans to spend $165 billion on three more U.S. plants, two advanced packaging facilities and an R&D center.
Wei said he would like to raise prices, but signaled TSMC would avoid the kind of abrupt increases seen in DRAM and SSD markets.
Can TSMC's $165B US bet overcome local talent shortages and rising costs to truly secure America's AI future?
As 'silicon inflation' grips the market, how will the surging cost of chips reshape the accessibility of future AI technology?
AI Chip Crunch 2026: TSMC’s Pivotal Role, Global Supply Chain Risks, and the Race for Advanced Packaging
Overview
As of mid-2026, the global technology industry faces a major bottleneck due to a shortage of advanced AI chips, driven by immense demand for leading-edge silicon. This shortage is severely impacting product development and rollout timelines, with no quick resolution in sight. The industry’s heavy reliance on TSMC’s production in Taiwan highlights the vulnerability of the supply chain, prompting the US government to push for more domestic capacity. TSMC’s 5nm and 3nm chips are central to its growth, but ongoing supply constraints mean that access to these advanced chips will continue to shape the pace of innovation and market competition.