Intel confirms work underway on 1.0nm and 0.7nm process nodes

Intel’s new CEO Lip-Bu Tan initially seemed to doubt efforts to turn around the company’s own fabs and warned that Intel’s 1.4nm process could be the last one, but the company’s stock is now ironically rising thanks to hopes placed in the fabs rather than the products, in part due to investor hopes for contracts with Elon Musk and Apple. And it now appears that Intel’s manufacturing process technology once again have a future.

This week CEO Lip-Bu Tan officially confirmed that the company is already working on two further generations of silicon technology following the 1.4nm process (designated Intel 14A): The first of these will be the Intel 10A process, meaning 1.0nm (“A” stands for ångström, and 1 Å = 0.1 nm). The second will be Intel 7A, which therefore translates to a 0.7nm technology. Or at least nominally. These figures are essentially arbitrary for all modern manufacturing processes and are intended to reflect relative progress versus previous generations, not the actual dimensions of structures inside the chip.

• Red more: Future Apple Silicon processors will be fabbed by Intel Foundry

The Intel 10A process already appeared in Intel documents during the Patrick Gelsinger era, however afterward its fate hung in the balance, as already mentioned, and it was therefore unclear whether it wouldn’t actually be canceled. With this confirmation, however, it appears that development ultimately received the green light.

In contrast, Intel is speaking publicly about the Intel 7A process for the very first time. It is of course logical that some roadmap for a successor to the 1nm process had to exist, but until now the company had not publicly discussed it or disclosed what name—or “nanometer number,” if you will—it would carry. Taiwanese foundry leader TSMC, which currently leads in manufacturing the most advanced chips, has so far confirmed only its 1.4nm process, while its 1.3nm and 1.2nm derivatives belong to the same family. It is fairly probable that afterward TSMC will also transition to a 1.0nm technology, but what step will follow after that remains unknown.

Both Intel processes are expected to use a new generation of EUV lithography with so-called High-NA (“High Numerical Aperture”) machines. Intel is planning to deploy these already with the 1.4nm process. High-NA technology is intended to significantly shrink structures on chips and thereby enable further scaling of silicon lithography. However, it also comes with a disadvantage—it achieves this by focusing the mask’s patterns onto a smaller area, and because of this it will not be possible to manufacture chips as large as on previous generations of process technology (meaning the only path forward will be to split chips into multiple chiplets).

• Read more: Intel announces 1.4nm process, first node with high-NA technology

In practice likely only after 2030

When exactly will chips actually begin using these technologies remains uncertain. Earlier roadmaps expected 1.0nm process production to begin roughly two years after the 1.4nm process, but since then the latter has already slipped behind schedule, and only this October is version 0.9 of the process design kit—the so-called PDK—expected to become available. Intel originally planned to begin risk production on the 14A process in early 2026, which evidently did not happen. If the 14A process slips into 2027 or 2028, then pilot production on Intel 10A may begin only in 2029 or 2030—and additional delays may occur there as well. We are therefore almost certainly talking about something that will arrive only in the next decade.

Sources: Tom’s Hardware

English translation and edit by Jozef Dudáš

⠀