Apple and Intel would dominate the 2nm chip race in 2026

The Taiwanese Semiconductor Manufacturing Company (TSMC) is planning to begin manufacturing chipsets using its 2nm process technology (N2) in late 2025 and set the target for delivering the first batch in early 2026. The N2 technology adopts the gate-all-around (GAA) transistor structure. It now looks like Apple and Intel might be the first customers for the 2nm chips.

In the last 20 years, TSMC's success was largely conditioned by its ability to offer a new manufacturing technology with power, performance and area improvements every year, while introducing a brand-new node every 18 to 24 months and maintaining high yields. However, as the complexity of modern fabrication processes are getting highly advanced and complex, it is getting much harder to keep the pace of innovation like before. With the N2 chip, there will be a strategic shift in TSMC's strategy of node development.

Apple has been TSMC's largest customer by revenue for the past decade. The company makes nearly all of Apple's chips including the ones used in iPhones, iPads and also the new M1 powered Macs. Apple's latest gadgets use a 5nm chip. The next generation of Apple devices are expected to use TSMC's 3nm chips, including the iPhone 14, to be announced later this year. However, we can only expect 2nm chips inside Apple devices after 2026.

Intel, on the other hand, will adopt the 2nm technology in its GPUs and other SoCs. Analysts from China Renaissance Securities speculate that Intel might use the N2 for the graphics tile of its next-gen client processor, code named as Lunar Lake. For the CPU tile in Lunar Lake, Intel will use its own 18A node.

Source: Computing.co.uk