Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

"Quantum Supremacy" expected at 50 qubits

  1. Sep 1, 2016 #1
    In an article posted here: http://arxiv.org/abs/1608.00263
    the authors estimate that a 50-qubit computer would be capable of computations that could not be replicated in real time by a classical computer.

    Interestingly, according to "New Scientists" (https://www.newscientist.com/articl...-googles-plan-for-quantum-computer-supremacy/), the authors of this article are Google engineers.

    My first reaction is to quibble with the term "supremacy" as it implies that this new technology will replace current classical computers. In fact, few of the chores assigned to current computers are appropriate for quantum processors.

    The article is also forecasting that Google will have a useful quantum processor by the end of next year.
  2. jcsd
  3. Sep 1, 2016 #2


    User Avatar
    Science Advisor
    Gold Member

    It is perhaps worth noting that whereas it is technically true that they are Google engineers; this is pretty recent. Most of them were "normal" academics until a couple of years ago when Google decided that they wanted to get serious about QC and essentially bought John Martinis group at UCSB.

    Martinis has been one of the leaders in he solid state QC field for a very long time (I'd say he is one of the top 3-4 people in the world) and has previously been in charge of big projects funded by e.g. iARPA; meaning he isn't just a random engineer at Google. Because of the money from Google he has also managed to hire even more good people and he also has access to some very good equipment.

    Hence, he is in a pretty good position to make predictions about the near future (I attended one of his talks a couple of months ago where he briefly discussed this)
  4. Sep 1, 2016 #3
    The D-wave machine is already at 512 bits (maybe more now), but does it perform Quantum computations I suppose is the question. I do question your statement about few chores being appropriate though. Any search algorithm or simulated annealing task is appropriate and we use plenty of those.
  5. Sep 2, 2016 #4
    The computation that the D-Wave machine performs is called "annealing", a term coined by D-Wave.
    It is a potentially useful function, but far from a general purpose quantum computer. It cannot support Shor's algorithm, Grover's algorithm, or anything other than "annealing".
  6. Sep 2, 2016 #5
    It's not a question of "tough", it's just a question of what can be expected of quantum computers. Yes, there are many applications for Quantum Computing, QC, but every one of them will require support from classical electronic data processing (CEDP).

    For example, suppose you want to liven up your Doom game by having those automated adversaries act in creative and diabolical ways. You will still need all of the existing CEDP Doom software. Plus you will need additional CEDP software to "drive" the QC processing. Plus you will need the QC processor(s) for the actual discovery of new methods for destroying the player's character with available resources.

    I would not call that a "supremacy" situation. None of that 3-D graphics rendering, user interface, game scoring and maintenance, etc. can be efficiently done in the QC. And the part that can be done by QC is new, specifically added to take advantage of the new QC technology.

    So what we have is a "complementary" situation.
  7. Sep 4, 2016 #6
    Simulated annealing has been around for some time and is not a term coined by D-Wave. We use it quite a lot so for us it quite common place, used by us for the first time in the late 80s. No the D-Wave is not general purpose and whether it is really quantum is up for debate it seems. I guess I was questioning your "few of the chores assigned to current computers are appropriate for quantum processors" statement; search is also a very common function and seems appropriate.

  8. Sep 5, 2016 #7
    The search provided by Grover's algorithm is very interesting and, unless you're the NSA, potentially the most useful of the qubit algorithms discovered so far. But it isn't good for storing a database or retrieving selecting records from a stored database. For example, if you have 10 million patient records and you want to find the one with the largest bill, it's hard to imagine Grover's algorithm being of practical use. The problem is that you would need to load that database into Grover's algorithm - and then hold it there until you were ready to perform a new search.

    As best I can tell, Grover's algorithm comes into its own when you need to select an element from a set which can be generated in the QM domain. So, for example, if you loaded up a data base of potential customers and then, in the QM domain, generated all possible salesman itineraries that covered all those sites (presuming such an algorithm exists - a likely assumption), then you could find the highest ranking (shortest path, cheapest path, etc) from among those possibilities.

    There exist methods for attacking the traveling salesman problem that can be shown to come well within 1% of the optimal solution. But with Grover's algorithm, you could make the ranking function more sophisticated. Perhaps including timely breaks each day for calling home, or getting the salesman back to the hotel every Wednesday night in time for his favorite TV show.
  9. Jan 15, 2017 #8
    From Ken Regan, arguably the most interesting insight I've seen so far:

    https://rjlipton.wordpress.com/2016/04/22/quantum-supremacy-and-complexity/ ... quote:

    Quantum supremacy has a stronger meaning than saying that nature is fundamentally quantum: it means that nature operates in concrete ways that cannot be emulated by non-quantum models. If factoring is not in latex.png —let alone randomized classical quadratic time—then nature can do something that our classical complexity models need incomparably longer computations to achieve. We like to say further that it means nature has a “notation” that escapes our current mathematical notation—insofar as we use objects like vectors and matrices that have roughly the same size as classical computations they describe, but swell to exponential size when we try to use them to describe quantum computations.
  10. Jan 15, 2017 #9
    You are using the term search in a very simple way. A search is a way of inverting a function i.e. y = f(x); given y find x. Nearly all computing functions can be reduced to a search and it is therefor a fairly fundamental and generally applicable operation.
  11. Mar 6, 2018 #10


    User Avatar
    Science Advisor

    So this didn't work out but there are recent news: Yesterday, Google announced Bristlecone, a 72 qubit quantum processor with which they plan to achieve quantum supremacy. They have a picture of an engineer installing the chip but they don't give a time frame or details about how far they have already come.
    Last edited: Mar 6, 2018
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted