As part of the recent Open Innovation Platform (OIP) 2024 event, Taiwanese semiconductor contract manufacturer TSMC, which is a leader in the field, spoke about its plans to develop advanced technological processes – 2nm and 1.6 nm.
At the end of next year, TSMC plans to launch the production of chips according to N2 technology standards, which will be the first in the 2nm class, and a year later it will be the turn to the more advanced N2P and N2X, and then to A16. Moreover, TSMC plans to master the last three technical processes by the end of 2026.
Technically, the N2, N2P, N2X, and A16 have a lot in common — they are all built on GAA transistors, and the N2 family features high-performance SHPMIM capacitors that reduce transistor resistance and, consequently, increase performance.
In turn, the 1.6-nm A16 process will be powered from the back of the BSPDN, which provides even higher performance, but it will not be without drawbacks in the form of additional heat generation, which will have to be reckoned with.
That is, future chips manufactured using the A16 process technology with BSPDN in its current form are most suitable for use in artificial intelligence accelerators that will be used in data centers.
At the same time, N2P will offer a noticeable increase in performance relative to the basic N2 process without additional difficulties — this technology is best suited for mobile processors and chips for entry-level computers. Finally, N2X will be optimal for high-performance CPUs.