Yes, NASA Did Help Achieve Quantum Supremacy (Update)
Quantum supremacy using a programmable superconducting processor, Nature
“Our Sycamore processor takes about 200 seconds to sample one instance of a quantum circuit a million times–our benchmarks currently indicate that the equivalent task for a state-of-the-art classical supercomputer would take approximately 10,000 years. This dramatic increase in speed compared to all known classical algorithms is an experimental realization of quantum supremacy for this specific computational task, heralding a much-anticipated computing paradigm.”
Quantum Supremacy Using a Programmable Superconducting Processor, NASA
“Here, we report using a processor with programmable superconducting qubits to create quantum states on 53 qubits … While our processor takes about 200 seconds to sample one instance of the quantum circuit 1 million times, a state-of-the-art supercomputer would require approximately 10,000 years to perform the equivalent task.”
Keith’s 9:00 am EDT note: A month ago I noted that a paper on this achievement was posted and then removed by NASA from NTRS. It had Eleanor G. Rieffel as an author – as does this Nature apaper – the titles are identical. Let’s see if NASA bothers to mention their role in all of this – or not. When I asked PAO about this a month ago they did not want to talk about it. I wonder if they even know it has been published. NASA has not emailed anything out nor has it posted this on PRNewswire.
Keith’s 10:05 am EDT update: NSF has issued a statement. Still nothing from NASA.
Keith’s 1:45 pm EDT update: I just got an email from someone at the NASA Ames Headquarters building – but not from Ames PAO. They informed me that a story had been posted. This is important stuff – but other than this email – I’d have never been alerted to this.
NASA and Google Achieve Quantum Supremacy, NASA Ames
“Achieving quantum supremacy means we’ve been able to do one thing faster, not everything faster,” said Eleanor Rieffel, co-author on the paper on this result, published today in Nature, and the Quantum Artificial Intelligence Laboratory Lead at Ames. “And even though that one thing isn’t terribly useful, that it has been done at all is groundbreaking.”
Keith’s 5:07 pm EDT update: NASA PAO finally sent the press release out at 5:07 pm EDT – half a day after everyone else.
Did NASA Ames Achieve Quantum Supremacy? (Update), earlier post
This is a world first, no? Why in the heck would NASA not want to talk about this? Especially with, ya know, China and all!
I wonder if they even know what it means!
That’s the thing with quantum computing, actually. Papers will point to spectacular results, but on further reading one learns that these conditions are so heavily bracketed that wide applicability seems limited.
Well-limited experimental conditions are part of good design. I can also see that even though papers show limited results, the results are truly stunning. The future is bright indeed.
Computer Science is my background and quantum computing is a little over hyped and this is just one example of why. Googles claim is misleading as the IBM statement in the link Don1024 posted states. If quantum computing ever gets to the point where it’s practical what the general public doesn’t realize is that a quantum computer will only out perform a classical computer for a small set of problems. Some even think they can’t work at all. See: https://www.quantamagazine….
So, don’t hold your breath.
This result is experimental evidence that quantum computers are real and do work. The sole point of this experiment, other than PR, was to demonstrate that quantum computers do work. That is the definition of “quantum supremacy” that is being used here—a computation that was done in fewer steps than a classical computer would require.
Now it is unable to solve real problem instances that we would care about, because error rates are way way too high for configurable quantum computers that have to retain entangled state for many, many rounds. So after this result the only still viable impossibility argument is that error rates will never go down. I could see this being argued on the short term, but configurable quantum computers with low error rates seem easy to construct if/when we have atomically precise manufacturing. So it’s just a matter of time.
I was referring to a practical machine.
I may be getting out of my depth here, but I believe the term you’re looking for is “Turing Complete.” That’s the standard for a truly general purpose computer. In the case of quantum computing, most of what I’ve seen is really hard-wired to only do one sort of calculation. It might do it with blinding speed, but it can’t be reprogrammed to do another sort of calculation.
But that’s not an inherently bad thing. I personally like the old, analog computers; you can rig up a capacitor to do integrals at the speed of light. And I’ve got some ideas for an analog min/max/average monitor which would be better (I think) than the digital electronics used for spacecraft telemetry and housekeeping.
There is no such thing as a “Turing complete” quantum computer. Such a thing cannot exist. The nature of quantum computing is that any instance solves very specific classes of problems. In fact we really shouldn’t talk about “quantum computers” generally, but rather “quantum computers that implement Grover’s algorithm” or “quantum computers for simulated annealing”, etc.
Now a distinction that IS valuable to make is whether a quantum computer can solve a configurable instance of a problem, which this one cannot. This computer only solves random instances of the specific problem class it is designed to solve. Which is as useless as it sounds. Error rates in quantum systems are still Many orders of magnitude too high to solve any but the most trivial instances of real world problems.
Yeah… no… at least according to IBM:
https://www.theregister.co….
Scott Aaronson has an interesting dicussion about this (Google vs. IBM’s claims) on his blog: https://www.scottaaronson.c…
They have achieved quantum SPENDING supremacy re: the costs of SLS.
Quantum computing is both an unprecedented advance in computing power that will revolutionize our world and a technological dead end with no practical application.
Cryptographers would disagree…
Yes, I think the articles rely on enthusiasm about magic answers without explaining how all the questions need to be rephrased or that is reprogrammed in nearly as difficult ways. The answer is 42, now to figure out the question.