3 Minutes
TSMC is facing a tight supply of 2nm wafers as AI-driven demand pushes capacity to the limit. The foundry has notified clients of planned price increases that will run for four consecutive years, starting in 2026 — and the first round is expected to kick in on New Year’s Day.
Why the squeeze matters
Capacity for TSMC’s most advanced nodes is essentially booked through the end of 2026. That backlog stems largely from a rush to produce chips optimized for artificial intelligence workloads, which drives up wafer demand, labor needs and capital spending. The result is a classic success problem: the company’s expertise and reliability have made it the go-to partner, but meeting that demand at scale will take time.
Industry analysts expect 2nm wafer prices to rise by a single-digit percentage in 2026, though estimates vary. Some research houses put the range for advanced-node increases between 3% and 10% over the coming year. Even TSMC’s 3nm node — already projected to reach capacity limits by 2026 — could see about a 3% bump as customers compete for limited slots.
Will customers back off? Unlikely. Despite higher costs and the option of Samsung’s 2nm GAA process, buyers are still racing to secure supply. That means tighter inventories and an uphill battle for anyone not already locked into early allocations.
Apple appears to have moved fastest: reports say it secured more than half of initial 2nm capacity for flagship chipsets like the A20 and A20 Pro. That leaves rivals such as Qualcomm and MediaTek choosing between smaller allocations or shifting to alternative process variants like TSMC’s N2P.
TSMC is trying to expand production — building three new facilities aimed at 2nm manufacturing — but fab construction and ramp-up cycles are lengthy. So even with new factories in the pipeline, widespread relief for the supply crunch won’t be immediate.
For device makers and chip designers, the takeaway is clear: expect higher foundry bills and plan production schedules and budgets accordingly. For the broader market, the squeeze underscores how AI demand is reshaping semiconductor economics — and why advanced-node capacity has become a strategic bottleneck.
Source: wccftech
Leave a Comment