Hacker News Reader

Comment by brown9-2 | Hacker News Reader

Do I understand correctly that the experiment itself has to do with simulating quantum circuits using a classical computer, and showing that a real quantum computer can produce the result much faster than a classical computer simulating a quantum one?

It’s hard to parse what the actual thing being computed is.

View on HN

That is my interpretation but I need a more detailed paper because everything is quite vague. Based on what I read, the claim isn't that great (yet). Everyone is gunning to be the historical first here.

"Hey, we have a new fluid flow simulation processor that reaches CFD supremacy over classical computers... we pour water across the surface of things and the results are indistinguishable to reality compared to the days, weeks, months it would take to simulate and still wouldn't reach the same resolution of accuracy on a classical computer."

It's not quite as bad as an analog water simulator made of water. They have built a programmable machine that can run any quantum problem it can load into its qubits. The problem is, it has so few qubits that it can only load tiny problems. The claim is that it can still beat classical computers on some subset of these tiny problems.

Is it correct that the claim “that it can still best classical computers on some subset of these tiny problems” based on classical computers simulating quantum circuitry to compute those problems?

I’m wondering if there is some other way the problem could be computed without having to simulate qubits.

A lot of work has gone into looking for classical algorithms for these sampling problems since the mid-2000s, and we have found some speedups, but everything still suggests that they are truly exponential on classical hardware. The classical algorithms for these problems don't involve modeling qubits directly, but they do involve sampling exponentially many states and averaging over the results, which is a lot like modeling qubits.

It sounds like they're computing the most probable eigenstate of the system. Doing so on a classical computer requires doing a bunch of inefficient N-body quantum mechanics with randomized interaction terms. Doing so on a quantum computer is basically just making a real-live quantum measurement.

By taking an ensemble of real-live measurements, you can infer what the most probable eigenstate is, which is the "output" of the program.

I'm confused as well. This is what their youtube video [1] claims, but to me it seems obvious that a computer simulating another computer would be slower.

[1] https://www.youtube.com/watch?v=-ZNEzzDcllU

For classical computers, one computer simulating another is only polynomially slower. The claim here is that any classical computer will be exponentially slower at simulating this problem. The experiment is just to show that the scaling is really polynomial vs exponential, and the uncertainty comes from the fact that we still can't fit very large N problems into current quantum computers. For small N, the two curves could look similar.

I think the goal here is to show that the quantum computer is not just a complicated classical computer. To do this, you design a problem that should be "easy" on your quantum computer but exceptionally hard (practically impossible) on a classical computer.

Compare this to, for example, D-Wave. As far as I understand their quantum computer they could show it was faster but it was "X times faster" rather than scaling in a completely different way.

That's what I can't quite get my head around. A computation is a physical process that produces a desired output. Any physical system can evolve very quickly to a state that is incredibly difficult to compute- the problem is to harness that system to make it produce something useful.

Reply on news.ycombinator.com