“Quantum supremacy using a programmable superconducting processor”: supplement
drive.google.comThanks for posting this document. I love the detailed description of the experimental setup. Waiting on bated breath for Scott Aaronson to post about this on his blog so I can know how excited to be.
(Not enough of a Quantum Information Theorist to reliably seperate hype from substance in such a complex field.)
The gist of it is that they've been able to simulate a very small number of random quantum gates faster than a classical-compute simulation of the same thing. It's not useful for anything, and it's at the level of things that D-Wave was doing for processor calibration over a decade ago.
Claiming this as "supremacy" is premature.
"Quantum supremacy" is a technical term that simply means demonstrating that a quantum computer can in fact perform some computation (something artificial like sampling from a certain kind of distribution) faster than a classical computer, not that the computation is useful for anything rightaway. This has not been achieved so far, but is expected to be around the corner. See this article from July: https://www.quantamagazine.org/quantum-supremacy-is-coming-h... and some of the better-informed comments on these other threads: https://news.ycombinator.com/item?id=21047804 https://news.ycombinator.com/item?id=21043659
Thanks for breaking it down for me. I figured if it was capable of general purpose quantum computation there'd have been more noise about it in the press.