Friday 16 October 2026
Night Edition
ZOTPAPER

News without the noise

ZOTPAPER
Business

TSMC chip capacity becomes fierce battleground as AI demand drives record quarter

The world's leading chipmaker cannot manufacture quickly enough, with competition for factory space threatening industry growth targets

By LineZotpaper
Published
Read Time2 min
Sources2 outlets
Taiwan Semiconductor Manufacturing Co. (TSMC) has posted a record profit for the sixth straight quarter, but the results highlight an intensifying struggle: the company cannot make chips fast enough to satisfy the artificial intelligence industry's insatiable demand. With customers including Nvidia, AMD, Apple, Google, Amazon, Microsoft, Meta and OpenAI all competing for limited factory space, analysts warn that capacity constraints could soon become a bottleneck for the entire AI sector.

TSMC reported its latest record quarter on Thursday, driven by strong demand for its advanced chipmaking technologies. The company expects fourth-quarter revenue to be 58% higher than a year earlier.

But the AI boom has created unrelenting pressure on TSMC's manufacturing lines. Last year, the company said it had the capacity to make more than 17 million 12-inch equivalent wafers. Customers want more.

"The competition is extremely fierce," said Jordan Klein, an analyst at Mizuho Securities. "It's a game of who can get the most leverage and supply."

TSMC North American CEO Sajiv Dalal told CNBC he receives at least 10 calls a day about wafer capacity.

The company produces the vast majority of the world's most advanced semiconductors, including graphics processors and central processors from Nvidia, AMD and Broadcom, plus custom in-house AI accelerators for an expanding roster of tech giants.

Klein warned that if compute demand continues to grow at the current rate over the next 12 to 18 months, capacity constraints at TSMC could threaten some of the industry's massive infrastructure investments and the "aggressive revenue growth targets" from AI companies.

§

Analysis

Why This Matters

  • The AI industry's ability to scale depends almost entirely on TSMC's manufacturing output; if the foundry cannot keep pace, product launches and service expansions could be delayed.
  • The capacity crunch could affect the returns on hundreds of billions of dollars in capital expenditure by hyperscale cloud providers and AI startups.
  • TSMC's pricing power and allocation decisions will shape which AI companies grow fastest, potentially concentrating market power among the largest customers.

Background

TSMC is the dominant manufacturer of the most advanced chips used in AI workloads, from Nvidia's GPUs to cloud giants' custom accelerators. As AI deployment has accelerated over the past two years, demand for TSMC's leading-edge nodes has outstripped supply. The company has been expanding capacity, including building new fabs in Arizona and Japan, but construction timelines mean new lines take years to come online. The latest earnings underscore both the financial benefit to TSMC and the structural tension it creates with customers.

Key Perspectives

AI companies (including hyperscalers and chip designers): They need guaranteed access to TSMC's advanced nodes to power their products and cloud services. The competition for allocation is intense, and those unable to secure enough capacity risk falling behind in the AI race. TSMC: It must balance allocating capacity across a growing customer base while maintaining its technological lead and investing in new fabs. The company benefits financially but faces operational and geopolitical pressure to expand. Analysts and investors: They see capacity as a potential chokehold on the entire AI sector. If constraints persist, revenue growth projections for AI firms could prove overly optimistic, and the cost of compute may rise.

What to Watch

  • TSMC's capital expenditure announcements and updates on new fab construction timelines.
  • Any public statements from major customers about supply constraints or in-sourcing chip production.
  • Q4 results and forward guidance from TSMC, particularly any changes in capacity outlook or allocation policy.

Sources

Zotpaper

Written by software from the reporting listed above, scored by an automated standards desk, and published without a person reading it first. If something here is wrong, tell the editor and it will be put right.

How we workSubscribe