IBM unveils world’s first 5nm chip

[Hogg]

Built with a new type of gate-all-around transistor, plus extreme ultraviolet lithography.

IBM, working with Samsung and GlobalFoundries, has unveiled the world's first 5nm silicon chip. Beyond the usual power, performance, and density improvement from moving to smaller transistors, the 5nm IBM chip is notable for being one of the first to use horizontal gate-all-around (GAA) transistors, and the first real use of extreme ultraviolet (EUV) lithography.

GAAFETs are the next evolution of tri-gate finFETs: finFETs, which are currently used for most 22nm-and-below chip designs, will probably run out of steam at around 7nm; GAAFETs may go all the way down to 3nm, especially when combined with EUV. No one really knows what comes after 3nm.

2D, 3D, and back to 2D

For the longest time, transistors were mostly fabricated by depositing layers of different materials on top of each other. As these planar 2D transistors got shorter and shorter (i.e. more transistors in the same space), it became increasingly hard to make transistors that actually perform well (i.e. fast switching, low leakage, reliable). Eventually, the channel got so small that the handful of remaining silicon atoms just couldn't ferry the electricity across the device quickly enough.

FinFETs solve this problem by moving into the third dimension: instead of the channel being a tiny little 2D patch of silicon, a 3D fin juts out from the substrate, allowing for a much larger volume of silicon. Transistors are still getting smaller, though, and the fins are getting thinner. Now chipmakers need to use another type of transistor that provides yet another stay of execution.
Enter GAAFETs, which are kind of 2D, but they build upon the expertise, machines, and techniques that were required for finFETs. There are a few ways of building GAAFETs, but in this case IBM/Samsung/GloFo are talking about horizontal devices. The easiest way to think of these lateral GAAFETs is to take a finFET and turn it through 90 degrees. Thus, instead of the channel being a vertical fin, the channel becomes a horizontal fin—or to put it another way, the fin is now a silicon nanowire (or nanosheet, depending on its width) stretched between the source and drain.

In the case of IBM's GAAFET, there are actually three nanosheets stacked on top of each other running between the source and drain, with the gate (the bit that turns the channel on and off) filling in all the gaps. As a result, there's a relatively large volume of gate and channel material—which is what makes the GAAFET reliable, high-performance, and better suited for scaling down even further.

Fabrication-wise, GAAFETs are particularly fascinating. Basically, you lay down some alternating stacks of silicon and silicon-germanium (SiGe). Then you carefully remove the SiGe with a new process called atomic layer etching (probably with an Applied Materials Selectra machine), leaving gaps between each of the silicon layers, which are now technically nanosheets. Finally, without letting those nanosheets droop, you fill those gaps with a high-κ gate metal. Filling the gaps is not easy, though IBM has seemingly managed it with atomic layer deposition (ALD) and the right chemistries.
One major advantage of IBM's 5nm GAAFETs is a significant reduction in patterning complexity. Ever since we crossed the 28nm node, chips have become increasingly expensive to manufacture, due to the added complexity of fabricating ever-smaller features at ever-increasing densities. Patterning is the multi-stage process where the layout of the chip—defining where the nanosheets and other components will eventually be built—is etched using a lithographic process. As features get smaller and more complex, more patterning stages are required, which drives up the cost and time of producing each wafer.