TSMC A10: The chip for the 2030 iPhone is already being prepared in Taiwan
In the silicon industry, the future is always planned years in advance. While TSMC has just begun volume production on the 2 nm process, the Taiwanese foundry is already looking toward an even more significant milestone: sub-1 nm, with a pilot production phase reportedly set for 2029.
TSMC: A roadmap that continues to tighten
This transition will not occur in a single leap. TSMC officially initiated volume production of its N2 node in the fourth quarter of 2025, confirming the shift to the 2 nm era. Concurrently, the company has unveiled its A14 node, expected in 2028, which promises performance gains of up to 15% at equivalent power consumption, or 25 to 30% energy savings at equivalent performance compared to N2.
In this context, a new report attributed to DigiTimes positions the next step: a process below 1 nm that TSMC is preparing for pilot production as early as 2029.
According to rumors, the company aims for an initial output of around 5,000 wafers per month and will rely on its Tainan A10 site as well as P1 to P4 facilities to facilitate this transition.
Why this step really matters
The sub-1 nm node is not just a battle of nomenclature. As nodes advance, the density of transistors, energy efficiency, and thermal control become crucial—especially for mobile devices and AI-related workloads. TSMC already presents its A14 as a technology designed to simultaneously enhance performance, consumption, and logical density, targeting AI accelerators, data center processors, and high-end mobile chips.
For Apple, the stakes are clear. The brand has traditionally been one of the first clients to leverage new TSMC processes in iPhones, iPads, and Macs. If the current trajectory holds, Apple chips at the turn of 2030 could experience a notable increase in density and efficiency when AI on-device requires more computing power without sacrificing battery life.
A technological advancement… that won’t be cheap
Nonetheless, extreme miniaturization does not solve everything. Even at 2 nm, yield rates remain a central concern, although TSMC previously indicated that the defect density of the N2 node is improving favorably compared to some predecessor nodes at similar stages. Additionally, manufacturing costs at the most advanced nodes remain very high, which typically limits these technologies to the most profitable market products.
In other words, sub-1 nm might initially benefit the most exclusive chips—premium smartphones, high-end machines, processors designed for intensive AI tasks—before gradually spreading to the rest of the market. This is often how the semiconductor economy works: innovation arrives first at the top before trickling down as volumes increase and costs stabilize.
More than a feat, an industrial imperative
What is at stake here goes beyond mere fascination with increasingly smaller numbers. As software becomes heavier, AI models make their way into devices, and battery life remains a non-negotiable requirement, advancements in lithography become a major strategic lever again. Therefore, sub-1 nm is not just an engineer’s promise; it is the condition for the next wave of high-end personal computing.
TSMC has yet to officially confirm this timeline for pilot sub-1 nm production in 2029. However, if this schedule holds true, the industry has already begun preparing for the next decade.
And as is often the case in silicon, the products that will inspire tomorrow are decided long before they reach our hands.



