TSMC is weighing a price increase of up to 20% for its CoWoS advanced packaging technology, according to a Trendforce report citing supply chain sources and local media. CoWoS, which combines logic dies and high-bandwidth memory (HBM) stacks in a single package, is already one of the tightest links in the AI supply chain. TSMC has not yet confirmed the report.
The pressure stems from escalating demand for AI accelerators. Nvidia's Blackwell chips pair two large GPU dies with eight HBM3E stacks, and upcoming Rubin and Rubin Ultra designs are expected to push integration further, increasing pressure on packaging and the supply chain. Testing has also become a bottleneck: Blackwell reportedly takes three to four times longer to test than Nvidia's previous-generation Hopper, while Rubin could require as much as seven times longer. Server CPUs and custom ASICs are also competing for packaging capacity.
Equipment lead times have reportedly stretched from three to six months to as long as eight months, prompting TSMC to place this year's equipment orders at existing prices rather than negotiating discounts or extra features, a departure from its usual approach. The company is also broadening its supply chain; earlier this month, GlobalFoundries agreed to produce silicon interposers for CoWoS in the U.S. under a five-year contract worth $2 billion, with production expected to ramp up in the first half of 2028. However, interposers are only one part of the packaging flow, so the agreement alone will not eliminate shortages.
TSMC is also reportedly considering broader price increases for 2027, with wafer prices on advanced process technologies expected to rise by 5% to 10%. Customers using 2nm and 3nm-class nodes could face the steepest hikes, and those seeking capacity beyond their original allocations may pay premiums of up to 25%. Industry estimates cited by Trendforce put TSMC's monthly capacity at over 106,000 wafers in 2026, with a target of 227,000 wafers in 2027.
The potential cost increases are likely to push up the price of AI hardware, with some of the burden expected to pass down the supply chain.