if you want a little more in depth explanation of the whole multiplication thing Ican recommend TAOCP volume 2. it has a section called 'How fast can we multiply?' it should provide more insight. there is a paper from djb (Daniel J Bernstein),which I can also recommemd: https://cr.yp.to/lineartime/multapps-20080515.pdf
from my basic understanding, memory is much easier to produce then logic chips like GPUs and CPUs, they don't need that many photolithographic layers. while it could be possible to produce memory in fabs for CPUs (though not really desirable in regard to costs) the other way round is more difficult
ORMs do have their use but you can easily screw things up. An anecdote from an university: their was a student administration system where students could themselves enroll to classes. simple enough job, one would guess. but there was a catch: at certain times, usually when more than one student logged in, the system predictably crashed.It turned out, that when a student logged in, a join over 13 tables was performed, even classes the student attended years ago where fetched at the login. These joins were clearly from misconfigured hibernate classes, took them some time to reduce the load on the system
I, too, have very fond memories of Mir Publishers. I was on vacation in Greece and had a stop in Athens, where I found by chance a bookshop which had books from Mir. the shop had irregular opening times but I managed to get there as often as I could, the books were dirt cheap
I'm still a little bit confused why the EU does not take action in this. This is definitely a monopolist overreach which has to be shutdown from the beginning