The pace at which leading foundries are advancing in their relentless pursuit to enhance the speed, performance, and efficiency of smartphone chips is truly astonishing. This year, TSMC, the foremost foundry in the world, is poised to commence mass production of 2nm chips.
Looking ahead to next year, the Taiwan-based company has announced plans to begin mass production of 1.6nm chips.
As the numbers of these process nodes decrease, the size of the transistors within the chips diminishes, allowing for a greater number of transistors to be packed into them. This is significant because smaller transistors enable a higher concentration in a specific area of the chip.
This measurement, referred to as transistor density, typically increases as the process node decreases. The overall transistor count of a chip is also crucial; generally, a higher number of transistors results in more powerful and energy-efficient semiconductors.
Take note of the remarkable reduction in process nodes observed over the past few years. For instance, in 2019, the iPhone 11 series utilized the 7nm A13 Bionic application processor (AP), which contained 8.5 billion transistors.
In September of this year, the iPhone 16 Pro Max was introduced, featuring the 3nm A18 Pro AP. Although Apple has not disclosed the transistor count for this chipset, it likely exceeds 20 billion transistors, especially since the A17 Pro has 19 billion. TSMC is experiencing significant growth, having reported a 37% year-over-year increase in fourth-quarter revenue, reaching $26.88 billion.
However, TSMC anticipates what it refers to as “smartphone seasonality” will result in a sequential decline in its revenue for Q1 2025, although on an annual basis, first-quarter gross revenue is projected to rise by 34.7%.
With its 2nm chip production, TSMC will implement Gate-All-Around (GAA) transistors that utilize vertically stacked horizontal nanosheets.
This design allows the gate to encompass all four sides of the channel, which helps to prevent current leaks and enhances the drive current. The outcome is chips that perform better while being more energy-efficient. When TSMC begins production of 1.6nm chips, it will introduce backside power delivery (BPD). BPD shifts power delivery from the front of a silicon wafer, where it limits space for transistors, to the back, where it is free from interference by other wires.
To illustrate our progress, consider that the original iPhone, launched in 2007, utilized a chip manufactured on the 90nm process node. The forthcoming iPhone 17 series, set to debut this September, will be equipped with the 3nm A19 and A19 Pro application processors (APs), which are produced using TSMC’s third-generation 3nm node (N3P).
Consequently, Apple is expected to introduce the first iPhone powered by 2nm silicon with the iPhone 18 series in 2026. As for when we can expect to see the first iPhone featuring an AP built on the 1.6nm node, we’ll need to provide an update on that later. In the meantime, TSMC reports that 1.6nm chips will deliver an 8% to 10% performance improvement at the same power level compared to the 2nm node.
