TSMC is aggressively scaling its 2 nm N2 node manufacturing, aiming for an ambitious output of 100,000 wafers per month by the close of 2026. This represents a fivefold increase in production capacity within approximately four months. Currently, the N2 node contributes a modest 3% to TSMC's revenue, significantly less than the 3 nm (30%) and 5 nm (33%) nodes. However, the projected growth suggests N2 could eventually account for over 10% of the company's revenue share.

Meanwhile, demand for TSMC's 3 nm N3 family of nodes remains robust. Major customers, including NVIDIA, AMD, and Broadcom, have driven N3's monthly wafer output to 180,000 units, surpassing TSMC's targets and prompting consideration of further expansion for this node. This surge in N3 demand has pushed its production capacity to its limits.

The N2 node, while in its early stages, shows strong promise. It has already achieved four times the number of tape-outs compared to the previous 3 nm N3 node. Leveraging GAAFET technology, N2 offers significant improvements, delivering up to a 15% performance boost at the same power level or a 30% power reduction at equivalent speeds. This technological advantage has attracted a growing customer base, including AMD for its EPYC server CPUs and Apple for its A20 Pro chips, indicating a clear trend towards N2 as the future direction for leading-edge semiconductor manufacturing.