Allen
Hotshot
Joined: Oct 1999
Posts: 8,839
Ohio USA
Originally Posted by Vaderini
...seems right no?
Yes, in my opinion (I do not know anything for sure -- as even expert engineering guesses about the future are at least partly wrong most of the time, literally -- and I'm no expert).
The second article states it. Very loosely, the "minimum" feature size is the node size. That is, if you have a 10nm node, things must be that large or larger. So, one manufacturer's 10nm node may actually have more transistors overall than another manufacturer's 7nm node. Intel makes much of their own stuff. So, they might get more out of a node than most.
Regarding the first article, I cannot read "Nederlands" (assuming that's what it is). However, in general, I think the smallest node possible is just a guess by almost everyone who opines (there must be at least two really smart folks on Earth who have already figured out the true minimum for Silicon CPUs). It may be possible for many features to be very small on a "simple" circuit, but not on a complex one (e.g. CPU). Also, we're getting down to a few "atoms" of distance with Silicon circuits -- realistically, that limits things.
All forgoing my opinions/guesses about the future -- that are virtually guaranteed to be wrong more than half the time