Wikipedia shows this history of the progressive miniaturisation of semiconductors:
10 µm – 1971
6 µm – 1974
3 µm – 1977
1.5 µm – 1981
1 µm – 1984
800 nm – 1987
600 nm – 1990
350 nm – 1993
250 nm – 1996
180 nm – 1999
130 nm – 2001
90 nm – 2003
65 nm – 2005
45 nm – 2007
32 nm – 2009
22 nm – 2012
14 nm – 2014
10 nm – 2016
7 nm – 2018
5 nm – 2020
Future
3 nm – ~2022
2 nm – >2023
Is there a physical limit on how small semiconductors could get, and if so, what is it?
Moore's law does not describe the size of semiconductors, but the size of transistors on a semiconducting chip. The chip itself remains a whole piece of semiconductor, several millimeters (millions of atoms) large, but the density of transistors (i.e., the elements of computer logic) on it increases.
In principle, a single atom (of size of about 0.1nm) could serve as a logic gate, although the necessary wiring, cooking etc. would require more place. What in reality is the crux of the Moore's law is that by now we are able to produce structures that are smaller than the electron coherence length (a few hundred nanometers), which means that we cannot use anymore the semi-classical theory if semiconductors, which served well for more than half a century, and which is grounded in boltzmann equation. Instead we need to use full quantum description - starting with Landauer-Büttiker formalism for non-interacting electrons, and so on.
