D-Wave says its chips use quantum mechanics to solve gargantuan problems.
What’s the News: Quantum computing is so complex an idea that even experts have a hard time telling whether a computer is actually “quantum.” But D-Wave Systems, a startup that’s made news and drawn skepticism over the last four years for claiming to have developed a quantum computer, has just made their first sale, to the defense contractor Lockheed Martin. And recent research shows that despite the suspicions D-Wave has endured, there may be at least something to their claim.
How the Heck:
- D-Wave calls their device, a 100-square-foot cube named D-Wave One, a “quantum optimizer.” That means that (at least in theory) it uses quantum mechanics to do one thing better than traditional computers: consider various different permutations of a problem simultaneously, thus coming up with an answer in a very short time. For problems where looking at all the options is the only way to find a solution, such as some kinds of code breaking, this capability is something of a holy grail.
- Lockheed says they plan to use D-Wave One to help solve their difficulties with melding hardware and software in defense technology, which, according to Technology Review, are the reason Lockheed’s F-35 strike fighter is more than 20% overbudget.
- The usual idea of a quantum computer is a machine that uses superposition–putting bits in more than one state at once–to solve such complex problems. The D-Wave system works along different lines: basically, engineers manipulate the states of quantum bits (qubits) to set up the problem, and if all goes well, the system should evolve into its lowest possible state, which represents the solution (see a great explanation of this at Ars Technica).
The D-Wave One system, pictured here with founder Geordie Rose.
What’s the Context:
- One reason physicists have been skeptical of D-Wave’s claims is that it’s not clear that the system is getting to that lowest state quantum mechanically; it could just as well be getting there through thermal fluctuations, which fall under the heading of classical physics. And if that’s the case, it’s not going to be faster at these kinds of problems than a normal computer.
- But in a recent Nature paper, D-Wave researchers reported that at extremely low temperatures, thermal fluctuations can be turned off, making quantum tunneling the only way their system has to get to the lowest state. That’s a step in the right direction, but critics have pointed out that further quantum mechanical effects would be required to actually make the computer more efficient than a traditional one, ScienceNOW reports.
The Future Holds: Plenty more work, both theoretical and practical, will be required on the part of D-Wave to convince the physics community that they’ve got a quantum machine. But this Lockheed sale (to the tune of $10 million for the device and a contract for keeping it running) adds a new twist. Since Lockheed will be undertaking the first serious, commercial use of the system, it will be interesting to see if it will share its experiences with outside scientists. Given the contractor’s reputation for secrecy, though, this experiment may remain as much of a black box as D-Wave one itself.
Image credit: D-Wave Systems
June 1st, 2011 at 2:39 pm
A volume measured in square-feet? What will quantum computers do next…
June 1st, 2011 at 4:21 pm
Have you heard the news? The internet is now on computers…
June 1st, 2011 at 6:11 pm
You know what I don’t get? This post.
June 2nd, 2011 at 4:40 am
Can it run Modern Warfare 3 when it comes out. ????
If so i’ll take two.
June 2nd, 2011 at 2:53 pm
If I had the wish to make money by encouraging the US military I’d buy Lockheed stocks right now. Too good I have the choice of not doing it! Think about it.
June 2nd, 2011 at 10:09 pm
100 square foot cube. Interesting that’s 100/6 =16.6667 per face = 4.08 sides = 68 cubic feet. Or did the author mean a 10 x10 face making a 1000 cubic foot cube? Maybe science authors should know the basics before they pen an article.
June 3rd, 2011 at 9:17 am
I think it’s exciting news! Lockheed surely put it’s people on to verify that the machine could do what they wanted. This is just a first step – much like the Cray-1 with it’s 80 Mhz processing capability. This is the kind of moment that passes without notice but then 30 years later you look back and see that it was a big part in how you arrived at where you are now…..
June 3rd, 2011 at 11:24 am
Maybe it’s going to ask what the costs would be for the Cyberdyne Systems Model 101…
June 4th, 2011 at 9:00 am
Hahahaha!, I saw that too Iain, 100 square foot cube? Wtf?
June 6th, 2011 at 12:46 am
What operating system are theyusing if its Linux that’s great but if its propreitary that’s not great and they will lose in the long run as so many high tech companies have and I worked for many and most are no longer in business.
June 6th, 2011 at 7:41 pm
If the picture is really Geordie Rose standing in front of the D-Wave, no edits, no forced perspective tricks, etc., then the following is reasonable:
The cube measures 10 x 10 x 10 feet, for a volume of 1000 cubic feet. However it’s footprint (floor space considerations in the data centre) is 10 x 10 feet. It’s a sloppy reference, and the height is higher than many data centres would routinely accomodate. However I can see how the vendor might market it by the floor footprint.
The near universal adoption of rackmount equipment in data centres has meant standarization of the physical space that servers, networking gear and storage take up.
July 4th, 2011 at 2:05 pm
Good. very great blog post!