Surfacing from being submerged by other responsibilities (including jury duty today), I wanted to point out two paper.
This preprint (based on a talk at the Solvay Conference - the 2022 one, not the 1927 one) from John Preskill provides a nice overview of quantum information, at a very accessible level for non-experts. It’s nice seeing this perspective.
This paper is much more technical and nitty-gritty. I’ve written before about my curiosity concerning how good the performance would be of quantum computing devices made with the process control and precision of modern state-of-the-art semiconductor fab. Here is a look at superconducting qubits made with a CMOS compatible process, using ebeam lithography and other steps developed for 300 mm wafer fabrication.
More soon….
What do you think of all the efforts to connect quantum information to holography and quantum gravity? Do you think it will end up being overhyped and a passing fad, or do you see it being a true permanent fixture in the field?
ReplyDeleteI find it rather fascinating how difficult it is to rigorously prove any exponential advantage of quantum computers over classical methods in the arena of condensed matter physics (and quantum chemistry more generally). I have to admit, I was initially under the very naive impression that the problem of high Tc cuprate superconductivity would be solved either by analog or digital quantum computers (e.g, ultracold atoms, or general purpose QCs). Now I've realized it's not clear whether they ever could do such a thing any better than normal classical computers.
ReplyDeleteHere is a very approachable lecture by Garnet Chan from Caltech on the topic. Quite remarkable to see we're still not even sure if an exponential quantum advantage for condensed matter exists! Lots of exciting problems for newcomers to work on :)
https://youtu.be/DZPH7ENcRLU
Anon@10:01, I do think there is a lot of hype in the presentation style, but there do seem to be some nontrivial connections between some of the ideas. The whole question of whether quantities like entanglement scale with volume vs area of boundary is interesting, though I have not had time to look deeply into this.
ReplyDeleteAnon@4:57, I was able to hear Garnet Chan’s talk on this at the March Meeting (see here). I do think analog quantum computers (I.e. simulators) can give us insights into model hamiltonians, their ground states, and their excitation spectra, and it seems like they should be able to do that more readily in some situations than trying to code up a classical simulation. Definitely a very interesting open question which problems in physics would benefit most from, say, general purpose quantum computers.
Since analog simulators are just special cases of general purpose QCs, it is also an open question whether they will have any exponential advantage over classical methods. I think this fact has been obscured so far because we haven't reached the regime of accessing the true ground states of e.g. the Fermion Hubbard model. Once we reach that regime, we will see if state preparation scales nicely or not with qubit number.
DeleteCertainly agree analog simulators are useful either way, but exponential speedup is still an unresolved question, as surprising as that seems to condensed matter folks.