TSMC Weighs 20% CoWoS Price Hikes as AI Bottlenecks Widen
Foundry quotes climbing up to 20% land on Nvidia and AMD accelerators as equipment lead times stretch to eight months, raising manufacturing costs across next-generation processor lines.
Taiwan Semiconductor Manufacturing Company tightened terms for artificial intelligence customers on Nov. 1, 2026, as foundry clients confronted proposed price hikes of 10% to 20% on Chip-on-Wafer-on-Substrate advanced packaging lines.
The price adjustments center on Chip-on-Wafer-on-Substrate technology, an advanced packaging system that integrates graphics processing units and High Bandwidth Memory onto a shared silicon interposer. Packaging services represent 16% to 22% of total hardware manufacturing expenses for top-tier artificial intelligence accelerators.
The documents behind this briefing are linked above. East Asia Brief produces its English text with AI assistance under human editorial review, and does not translate or republish other outlets' articles. See our methodology and AI policy. Spotted an error? Tell us.
