>>12238954Well like you say, many-body physics is an interesting problem, and we've know since the 90s that Hamiltonian simulation is in general BQP-complete. We have decades of classical computational theory and experience to give us a good handle at when they are not feasibly simulated, e.g., for highly correlated systems. Thus it's pretty clear that this is a case where quantum has the advantage complexity-wise.
Quite frankly, a proper, functioning quantum computer would be a feat in and of itself. It's not really easy to answer your question because it's a question about "practicality", yet does not address what is meant by "functioning quantum computer". Error-corrected and fault-tolerant, sure, but there's still much more to it than that. A fault-tolerant QC with 1 T gate factory honestly might still take too long to perform any sort of interesting calculation, even if we don't have to worry about noise and error anymore. But one with a plethora of these factories would be an absolute fucking monster. Imagine if you could just take some textbook quantum circuit and throw it into the QC without any sort of compilation, and it just runs through every non-Clifford operation without hitch. That's about as close to sci-fi computation as you can get in the real world. But I also highly doubt that we will ever have devices like that -- it's so technologically advanced I'm not sure if we'll ever be able to just have magic states on-demand like that.
