The first smartphone chipsets mass-produced on TSMC’s 2nm node are expected to launch next month. That is only a small taste of how popular the Taiwanese outfit’s next-generation process has become.
At the 2026 Japan Technology Symposium, TSMC senior vice-president Kevin Zhang said 2nm had four times as many tape-outs as 3nm. This shows how fast chip designers are piling into the advanced node.
The 2nm node’s popularity is good news for TSMC and AI firms. It is less cheerful for the Fruity Cargo Cult Apple and everyone else stuck fighting for wafer scraps.
The first version of TSMC’s 2nm family, N2, entered volume production in the fourth quarter of 2025. After two more quarters, it contributed three per cent of the company’s wafer revenue, despite no commercial product having shipped.
Google’s Tensor G6 is tipped to become the smartphone industry’s first 2nm system-on-chip. It is expected to be followed by the A20 Pro in Job’s Mob’s upcoming iPhone 18 Pro and iPhone 18 Pro Max.
TSMC is riding a demand wave that makes the 3nm scramble look almost civilised. The N2 node will eventually be followed by N2P, a slightly quicker version offering a five per cent performance improvement.
N2P is expected to turn up in Qualcomm’s Snapdragon 8 Elite Gen 6 and Snapdragon 8 Elite Gen 6 Pro. MediaTek’s Dimensity 9600 is tipped for the same process.
The idea is to gain a small edge over Job’s Mob’s A20 Pro. It may only be a sliver, but chip marketing departments have built entire PowerPoint shrines from less.
N2X is planned for 2027 and will target high-end applications. DigiTimes reports that N2U is expected to enter mass production in 2028 and serve mainstream products.
Smartphone companies still account for a smaller share of TSMC’s overall 2nm story. The bigger beasts are likely to be AI GPU makers such as Nvidia, which will want as much capacity as possible.
That creates an awkward problem for TSMC. The same kind of supply choke that dogged 3nm could return, only with AI firms waving fatter chequebooks.
Job’s Mob may have to abandon 2nm after just two generations and move to 1.4nm to dodge shortage risks. That sounds tidy on a roadmap and rather less tidy in real fabs.
The remaining smartphone brands should not expect much sympathy either. TSMC’s priorities have been drifting towards non-smartphone customers because AI margins speak louder than handset whining.
