Stories
Slash Boxes
Comments

SoylentNews is people

SoylentNews is powered by your submissions, so send in your scoop. Only 10 submissions in the queue.
posted by Fnord666 on Tuesday December 08 2020, @09:26AM   Printer-friendly
from the seeing-the-light dept.

Quantum device performs 2.6 billion years of computation in 4 minutes:

I am a great believer in solving problems with lasers. Are you suffering from a severely polarized society and a fast-growing population living below the poverty line? Well, I have the laser to solve all your problems.

OK, maybe not. But when it comes to quantum computing, I am of the belief that lasers are the future. I suspect that the current architectures are akin to the Colossus or the ENIAC: they are breakthroughs in their own right, but they are not the future. My admittedly biased opinion is that the future is optical. A new paper provides my opinion some support, demonstrating solutions to a mind-boggling 1030 problem space using a quantum optical system. Unfortunately, the support is a little more limited than I'd like, as it is a rather limited breakthrough.

[...] Unlike both of these options, an optical quantum computer could be a (large) chip-scale device that is powered by an array of laser diodes, with read out done by a series of single-photon detectors. None of these requires ultralow temperatures or vacuum (if photon-counting detectors are required, then liquid nitrogen would be required). Optical quantum computing will require temperature stability and, as this paper demonstrates, a rather complicated feedback system to ensure that the lasers are working exactly as required. However, all of that could be contained in one large rack-mounted box. And that is, for me, the critical advantage of optical systems.

This does not mean that light will win though. After all, germanium is a better semiconductor than silicon, but silicon still rules the roost.


Original Submission

This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(1)
  • (Score: 0, Disagree) by Anonymous Coward on Tuesday December 08 2020, @09:47AM (5 children)

    by Anonymous Coward on Tuesday December 08 2020, @09:47AM (#1085168)

    On the other hand, the work is no different to other quantum-advantage experiments: take a problem that is mostly useless but happens to map exactly to the architecture of your computer. Naturally, the computer can solve it. But the point of a computer—and this is why the researchers do not refer to the device as a computer—is to solve many different useful problems. And for these cases, we have not yet seen undisputed evidence of the promised quantum advantage. I have no doubt that it will come, though.

    • (Score: 0) by Anonymous Coward on Tuesday December 08 2020, @11:08AM (1 child)

      by Anonymous Coward on Tuesday December 08 2020, @11:08AM (#1085174)

      Sorry, your post isn't sufficiently indicated as special by abuse of the markup. You need to use quote block and Courier font both.

      • (Score: 0) by Anonymous Coward on Tuesday December 08 2020, @11:20AM

        by Anonymous Coward on Tuesday December 08 2020, @11:20AM (#1085176)

        The post is entangled. Try ugain.

    • (Score: 2) by driverless on Tuesday December 08 2020, @12:16PM (2 children)

      by driverless (4770) on Tuesday December 08 2020, @12:16PM (#1085180)

      "Fancy physics device performs task specifically chosen to be easily solvable by fancy physics device but otherwise useless, inventors add the word 'quantum' and publish a press release" would be a better title.

      • (Score: 0) by Anonymous Coward on Tuesday December 08 2020, @03:12PM

        by Anonymous Coward on Tuesday December 08 2020, @03:12PM (#1085203)

        Cold Fusion here, get your cold fusion!

        Oh, wrong century? Sorry.

      • (Score: 0) by Anonymous Coward on Wednesday December 09 2020, @05:24PM

        by Anonymous Coward on Wednesday December 09 2020, @05:24PM (#1085608)

        Isn't a calculator just a fancy physics device? Isn't a processor one? Or a GPU? Isn't an FPGA or an ASIC the same thing? Or our brains? I mean, they all operate within the laws of physics, no?

        Doesn't an ASIC simply use fancier, more specific, physics to do tasks that a CPU would take forever to do at a much higher speed? So these people are just creating special use cases of a more specialized processor ... perhaps they are just creating different ASIC's.

  • (Score: 1, Insightful) by Anonymous Coward on Tuesday December 08 2020, @10:07AM (2 children)

    by Anonymous Coward on Tuesday December 08 2020, @10:07AM (#1085170)

    This computer does compute the result to a class of problems that is hard due to quantum effects, but that is it. It is similar to constructing a machine that can do precise measurements of different sized and types of pendulums with no friction and then running a weighted, quadruple pendulum experiment the in it. That machine will come up with the solution to the problem faster than a general purpose computer will due to the complexity involved. But, like TFA's computer, it is hardly generally useful, isn't what most people think of as a quantum computer (no qubits in the traditional sense), and won't be doing the quantum computing "magic" (things like Shor's algorithm) literally ever.

    • (Score: 0) by Anonymous Coward on Tuesday December 08 2020, @10:43AM (1 child)

      by Anonymous Coward on Tuesday December 08 2020, @10:43AM (#1085173)

      But can it run Crysis?

      • (Score: 0) by Anonymous Coward on Tuesday December 08 2020, @11:42AM

        by Anonymous Coward on Tuesday December 08 2020, @11:42AM (#1085178)

        No. Use your dual socket 128-core Epyc instead.

  • (Score: 0) by Anonymous Coward on Tuesday December 08 2020, @11:15AM (1 child)

    by Anonymous Coward on Tuesday December 08 2020, @11:15AM (#1085175)

    The ENIAC wasn't just a breakthrough: It was a Turing-complete, general-purpose, programmable, electronic computer. All of today's computers are descended from it. There were other early computers that shared some of those traits but not others. The Zuse machines were Turing-complete, but they were mechanical. The Colossus was electronic, but wasn't general-purpose and was barely programmable. Still, all of them, in some sense, did become the future.

  • (Score: 4, Interesting) by bzipitidoo on Tuesday December 08 2020, @11:37AM (12 children)

    by bzipitidoo (4388) on Tuesday December 08 2020, @11:37AM (#1085177) Journal

    There are analog computers that can do basic calculus such as integration of a curve. But that's all they can do. And we hardly use them even for that, because they can't integrate just any curve, the entire integral has to be within the range of the voltage the computer can produce. Yeah, sure, the analog integrator simultaneously computes the integral at every point of the curve, but this ability has not proven to be of much value.

    In contrast, neural net computation is much more flexible. We know neural nets can be trained to do different tasks, and produce useful results.

    The author has "no doubt"? Mathematicians were once confident that math could be "complete". All that they had to do was figure a way to show it. I wouldn't discount the possibility that quantum computation may yet prove to be impossible to harness, or so limited as to be nigh useless.

    I'd put my money on the P=NP question being proven first, before the development of useful quantum computers. The problems could well be intimately connected. A proof that P!=NP could be based upon a proof that general purpose quantum computing is not possible.

    • (Score: 2) by bradley13 on Tuesday December 08 2020, @12:09PM (5 children)

      by bradley13 (3053) on Tuesday December 08 2020, @12:09PM (#1085179) Homepage Journal

      Exactly this. So far, "quantum supremacy" has involved cooking up very specific problems that no one particularly needs solved, and showing that - in theory, at least - they could be solved with a quantum computer. As you point out, analog computers exist, and for certain applications are theoretically very powerful. And they haven't been taken seriously since - what - the 1950s? The applications are just too specialized, and the output just too imprecise.

      It's looking more and more like we can put quantum computing in the same category as fusion and true AI: Always just around the corner, practical applications just a few years away.

      --
      Everyone is somebody else's weirdo.
      • (Score: 3, Interesting) by takyon on Tuesday December 08 2020, @12:33PM

        by takyon (881) <takyonNO@SPAMsoylentnews.org> on Tuesday December 08 2020, @12:33PM (#1085182) Journal

        If someone figures out how to make working quantum chips on the same silicon wafers used to make normal CPUs, then it could go from nothing to tiers beyond "quantum supremacy" very quickly. Don't hold your breath though.

        --
        [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
      • (Score: 0) by Anonymous Coward on Tuesday December 08 2020, @01:34PM (3 children)

        by Anonymous Coward on Tuesday December 08 2020, @01:34PM (#1085186)

        Specialized maybe, but hardly useless: Just about everyone wants a quantum computer that can break encryption (and just about everyone also wants everyone else to not have it). Still, it's either possible or it isn't; we need to know. Right now it seems like an engineering problem, which is in the process of being solved.

        There are other useful problems besides encryption that quantum computers could help with, encryption is just the most widely known. Physics simulation would be helped tremendously, which could lead to all kinds of advances: protein folding, superconductors, who knows. There is lots of potential there. And the raw mathematics that quantum computers are supposed to make easy, factorization and discrete logarithms, might turn out to lead to new mathematical advances if they were easy to compute. The thing about math, it's just about impossible to predict what benefits it will bring down the road.

        No, quantum computers won't help you take better selfies or get faster framerates in the latest games, but there is more to computing than that. Maybe quantum computers are only ever useful in research or specialized business fields, but that's still useful.

        • (Score: 3, Funny) by takyon on Tuesday December 08 2020, @02:02PM (2 children)

          by takyon (881) <takyonNO@SPAMsoylentnews.org> on Tuesday December 08 2020, @02:02PM (#1085190) Journal

          Humanity needs 1 mm2 room-temperature quantum processing units in smartphones.

          --
          [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
          • (Score: 0) by Anonymous Coward on Tuesday December 08 2020, @07:00PM (1 child)

            by Anonymous Coward on Tuesday December 08 2020, @07:00PM (#1085280)

            ...in order to...

            (i) take selfies
            (ii) watch cat videos
            (iii) trade stocks

    • (Score: 4, Insightful) by Thexalon on Tuesday December 08 2020, @03:58PM (3 children)

      by Thexalon (636) on Tuesday December 08 2020, @03:58PM (#1085223)

      The point of NP-complete problems is that you can easily check whether you got the right answer, but can't easily develop the right answer from scratch. And yes, some of those are very specialized, e.g. Satisfiability or the TSP.

      Some of them are a lot bigger deals, though. For instance: "Given these axioms and this list of definitions of terms, prove this mathematical conjecture."

      --
      "Think of how stupid the average person is. Then realize half of 'em are stupider than that." - George Carlin
      • (Score: 2) by bzipitidoo on Tuesday December 08 2020, @06:34PM (2 children)

        by bzipitidoo (4388) on Tuesday December 08 2020, @06:34PM (#1085261) Journal

        Well, yes, NP stands for "Nondeterministic Polynomial", meaning that at every fork in the road, if we didn't have to choose (determine) which way to take, could take both ways, then these problems could be solved in polynomial time. Quantum superposition just might be a means of having it both ways, so to speak. And, yes, if you have been given a list of which branches to take at all the forks, you can quickly try it.

        Of critical importance is that these problems are all related. We can quickly (polynomial time) transform an instance of one NP problem into an instance of a different NP problem. That's basically how almost every NP problem has been proven to be NP: show how to transform 3-SAT instances into instances of the problem in question.

        • (Score: 1, Insightful) by Anonymous Coward on Tuesday December 08 2020, @06:53PM (1 child)

          by Anonymous Coward on Tuesday December 08 2020, @06:53PM (#1085278)

          Quantum computers don't solve NP-complete problems. The class of problems they can solve is called BQP, which exceeds P (obviously, or they'd be useless) but is still a subset of NP. Solving NP-complete problems would make them vastly more useful. They are still probably useful, but if they could solve NP-complete problems, they'd be revolutionary instead of just helpful in some specific cases.

          • (Score: 3, Interesting) by bzipitidoo on Wednesday December 09 2020, @12:25AM

            by bzipitidoo (4388) on Wednesday December 09 2020, @12:25AM (#1085361) Journal

            It is unknown if BQP is a proper subset of NP. If P=NP, then of course BQP=NP. If P!=NP, then it could still be that BQP=NP. And in those cases, quantum computers could solve NP problems in P time.

            If BQP!=NP, or BQP!=P, then also P!=NP. If you can prove either of those, then you should publish and if everyone is convinced you are correct, collect your $1 million prize.

    • (Score: 2) by HiThere on Tuesday December 08 2020, @05:11PM (1 child)

      by HiThere (866) on Tuesday December 08 2020, @05:11PM (#1085234) Journal

      Neural nets, though, are quite tricky to train. Train them too much, and you get "overtraining" which produces bad results. Train them on a subset of the data, and they may only learn to handle that subset. Etc. And it's quite difficult to know just what they're learning. See https://www.csail.mit.edu/news/why-did-my-classifier-just-mistake-turtle-rifle [mit.edu] .

      --
      Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
      • (Score: 0) by Anonymous Coward on Tuesday December 08 2020, @07:02PM

        by Anonymous Coward on Tuesday December 08 2020, @07:02PM (#1085281)

        They are like primitive microbes growing in a closed environment (typically manure-based).

  • (Score: 3, Informative) by Hartree on Tuesday December 08 2020, @02:46PM

    by Hartree (195) on Tuesday December 08 2020, @02:46PM (#1085199)

    "Well, I have the laser to solve all your problems."

    Nonsense. Everyone knows that lasers can only solve the world's problems when properly mounted on sharks.

  • (Score: 0, Funny) by Anonymous Coward on Tuesday December 08 2020, @05:51PM (1 child)

    by Anonymous Coward on Tuesday December 08 2020, @05:51PM (#1085241)

    And the answer was 42.

    • (Score: 1, Informative) by Anonymous Coward on Tuesday December 08 2020, @07:04PM

      by Anonymous Coward on Tuesday December 08 2020, @07:04PM (#1085283)

      -1 obvious

      So 41.

  • (Score: 0) by Anonymous Coward on Thursday December 10 2020, @12:06AM

    by Anonymous Coward on Thursday December 10 2020, @12:06AM (#1085770)

    Doing 2.6 billion years of computation in few minutes SOUNDS impressive, until you hear about the ship that made the Kessel run in less than 12 parsecs!

(1)