Scott has again picked up the mantle of Dwave critic. From New Scientist today, Aaronson disagrees. “I don’t think this is going to be directly used for anything relevant to Google’s business,” he says. “Google is a very wealthy company and they do have money to throw around on things that just might be fun to play with.”
Even if D-Wave managed to build (say) a coherent 1,024-qubit machine satisfying all of its design specs, it’s not obvious it would outperform a classical computer on any problem of practical interest. This is true both because of the inherent limitations of the adiabatic algorithm, and because of specific concerns about the Ising spin graph problem. On the other hand, it’s also not obvious that such a machine wouldn’t outperform a classical computer on some practical problems. The experiment would be an interesting one! Of course, this uncertainty — combined with the more immediate uncertainties about whether D-Wave can build such a machine at all, and indeed, about whether they can even produce two-qubit entanglement — also means that any talk of “lining up customers” is comically premature.
Now Aaronson concedes that Dwave has achieved entanglement with a quantum annealing system for its full 512 qubits.
USC researchers compared a D-Wave One device, which has 108 qubits, with two different high-end conventional processors running simulation software that has been optimised to perform the computations in the same way as D-Wave. In this case, although the D-Wave One showed some evidence of quantum behavior, it took longer – 15 microseconds – to solve a problem than the conventional processors, which took 4 and 0.8 microseconds.
Lidar is now conducting similar tests with the upgraded D-Wave Two, which has 503 qubits and so might perform better on larger problem that ordinary, or classical, computers struggle with. “The really interesting question is whether a quantum processor exhibits better scaling with problem size than all classical processors it is compared with,” says Lidar. “If this is the case, then it validates the promise of quantum computing.”
Google has performed its own speed tests of the Dwave 512 qubit quantum system versus classical quadcore workstations and finds Dwave to be 11,000 to 50,000 times faster on problems of interest in Google.
There were tests by Catherine McGeoch, author of A Guide to Experimental Algorithmics, has 25 years of experience setting up experiments to test various facets of computing speed, and is one of the founders of “experimental algorithmics,” which she jokingly calls an “oddball niche” of computer science.
They found the 439 qubit system up to 3600 times faster than a quadcore computer and a 502 qubit system to be up to 10,000 times faster.
Aaronson thinks that the classical algorithms and systems needed to be tuned and optimized.
Dwave has sold two systems. One to Lockheed (for ten million dollars in 2010) and one to Google (in 2013).
Google and Lockheed are indicating superior performance on the Dwave 512 qubit system for problems of interest to them.
Aaronson theory is that Google is being wild with their money. Pissing it away on what is not actually a better or more useful system.
There will be a quantum computer with over 100 qubits of processing capability sold either as a hardware system or whose use is made available as a commercial service by Dec 31, 2010. Predictor – brian wang