I'm sure great things are happening in the field, and many wonderful ideas and discoveries have been made — but statements such as "take us past Moores law" requires manufacturing, and I haven't seen anything to say that this is possible albeit I am ever hopeful for the day that it is.
[1] https://www.nature.com/articles/ncomms10177
Edit: spelling
The edge of a graphene sheet is a defect that affects the electronic properties, with strong orientation dependence
However, in this article they intentionally introduce additional defects to tune the electronic properties, so there’s probably no fundamental issue. I agree with another poster here that manufacturing at scale in integrated devices is not obviously viable. Especially because it seems like a different method of introducing defects is needed if you want to grow the graphene by chemical vapor deposition.
I thought Aluminium and titanium were used in the lower interconnect layers, but according to [1] you are right.
The tech mentioned in the article is supposedly useful for transistors, not necessarily interconnect though. I guess that would be useful for going beyond planar processes. But at some point our current architectures are already severely limited regarding TDP anyway, so unless you can afford using only 10% of your chips, an architecture revolution is overdue IMO.
[1] https://en.wikipedia.org/wiki/Interconnect_(integrated_circu...
As in: Moore's law is still giving you more transistors, but you can't actually turn them all on, and they aren't getting faster, so no free lunch.
(Which of course is why the Apple M1 is so significantly ahead, at least of Intel)
What about accessibility of the computers required to make it 1000x harder?
