The world's largest contract chipmaker is sounding the alarm: the AI boom is straining its factories to the limit, and the cost is likely to be passed on to everyone downstream.
TSMC CEO C.C. Wei stated flatly that it will be a "long time before we can meet customer demand," as demand for AI chips continues to far outpace what the company can produce. Even after TSMC's 3nm wafer production hit a new monthly milestone of 175,000 units, supply remains constrained for AI customers, according to Wccftech.
The price pressure is coming from multiple directions. In an interview with the BBC, TSMC's CFO said that further price increases for its foundry services seem "almost inevitable" due to inflation, compounded by shareholder pressure as the company pushes its facilities to capacity. One source, CryptoBriefing, cited signals of potential hikes in the range of 5 to 15 percent on advanced chips.
To close the gap, TSMC is mounting a massive global expansion—simultaneously building out manufacturing capacity in the United States, Germany, and Japan, while continuing to pour investment into its home base in Taiwan. The infrastructure spending is enormous, but executives make clear it won't solve the crunch anytime soon.
For the broader AI industry, this matters enormously: TSMC makes the chips inside products from Apple, Nvidia, AMD, and virtually every other major tech company, meaning any price increase ripples outward to data centers, consumer devices, and the cloud services that millions of people rely on every day.