TSMC has unveiled a four-year roadmap for price increases, driven by the unprecedented surge in demand for AI-related hardware. Production capacity for the upcoming 2nm nodes is reportedly fully committed through the end of 2026, with clients already notified of annual price adjustments set to take effect on January 1, 2026.
For cutting-edge technologies, including the 2nm process, price hikes are projected to range between 3% and 10%. Meanwhile, the 3nm node—the current industry benchmark for performance—is expected to see a 3% to 5% increase due to ongoing supply shortages. Leveraging its near-monopolistic market position, TSMC maintains the leverage to dictate terms and implement these cost escalations with minimal resistance.
Despite Samsung’s advancements with 2nm GAA technology, industry titans like Apple, NVIDIA, and AMD prefer to pay a “reliability premium” to TSMC rather than risk transitioning to competitors. Apple has already secured over half of the initial 2nm output for its future A20 and A20 Pro chips. Other major players, such as Qualcomm and MediaTek, may have to settle for remaining allocations or wait for future production windows.
These rising costs are attributed not only to skyrocketing demand but also to a global shortage of specialized talent and the massive capital investment required for infrastructure. While TSMC is currently constructing three new fabrication plants dedicated to 2nm production, the market is unlikely to see more affordable high-performance silicon until these facilities reach full operational capacity.
Source: iXBT.games