Google Just Bought A Quantum Computer! Or Did They? Or Do I Care?

Funny commentary on Google's new quantum computer, built by D-Wave

http://www.funtelligence.net/2013/07/google-just-bought-quantum-computer-or.html

New post from the guys at Funtelligence explains Google's new machine in a way that a drunk can understand. photo credit: jurvetson via photopin cc

Via photo credit: jurvetson via photopin cc

OK, I know that quantum mechanics is not something new. I also know that if a "true quantum" computer was actually built, it will likely have an Achilles heel because at some point, their speed still depends on conventional components. Put another way, they're like Jon Hamm without the paparazzi's favorite trouser snake; you just don't reach that critical mass necessary to make something commercially viable. So why would Google buy one?

Well, as Wired Magazine's Cade Metz reports in his article 'Google's Quantum Computer Proven to be the Real Thing (Almost)', just how quantum the computer is might be in dispute, but there's no doubt that manufacturer D-Wave has built a remarkable machine.

Before we go further, let's all get a refresh on what a quantum computer is.

I'm going to stop you in your tracks if you think that I am going to write some brilliant, accolade worthy post that explains this neatly and concisely. To my knowledge, no one has. Nope, this is it in a nutshell, followed by a lame analogy:

Contemporary computers are bound by the laws of regular physics, meaning that each physical component can only do only one thing at a time. Quantum computers can multi-task exponentially because they are governed by the laws of things that are so small (electrons and such) and moving so fast that they are only observable in terms of probabilities. That's actually still a little off because they don't have mass but my name isn't Higgs so I'm not going down that rabbit hole. By using quantum probabilities, you aren't tied down by making computations, you just let sub-atomic particles do what they do naturally.

Think of it this way - you own a bar and you've worked in that bar every night for twenty years. You know from experience that 99.99 percent of the time, the guy who pukes and falls off his stool will wind up pissing his pants and passing out right on the floor. You also know that there is a .01 percent chance that he will miraculously rally, have what Dave Chapelle dubbed "a great moment in hook-up history", and take a decent looking chick home because you've seen enough in 20 years to know that anything can happen. The 21 year old bar tender you hired a year ago has seen plenty of people puke and fall off the stool, but he's never seen anyone achieve a come from behind victory of that magnitude. To him, the chances of someone wiping the puke off and then throwing down the world's greatest one-liner on the hottie in the booth with all her friends is exactly zero. So, in this example, the bar owner has the quantum vantage point because he knows that sooner or later there's bound to be an outlier who will score, no matter how unlikely. The bar tender has the vantage point of classic physics in that every time he's observed someone puke and fall off a stool, it's time to wheel out the mop and call a cab.

The trick with computers is that you don't want to wait around and observe how often something will happen, which is basically the same thing as having a conventional computer physically run through every possibility, one at a time. They simply can't account for all the combinations that might happen on any given night. So, what D-Wave does is bring the temperature of their machine down to near absolute zero, all but freezing all the components (presumably just electrons bouncing about inside circuits they've made) in place. Then, the user can give the computer all of the relevant data, which is basically any possible factor that can influence the outcome (looks of the guy, number of trollops in the bar on any given night) that will be generated. Then, they heat the computer back up and let those electrons bounce around so fast that they can replicate 20 years worth of observations in a couple seconds and presto - anyone can have the perspective of a seasoned bar man!

Or, if you want an actual explanation, Wired also reported the following in February in Metz's 'D-Wave Defies Critics with World's First 'Quantum Cloud',:

[D-Wave’s computer is designed to solve what are called combinatorial optimization problems. The classic example is figuring out the most efficient route for a traveling salesman going to multiple destinations. There’s no mathematical shortcut that computers can take to solve combinatorial optimization problems. They have to use brute force: Simply check all possible combinations. The trouble is, the number of possibilities explodes exponentially with the problem size. For example, if you have six destinations, there are 64 possible combinations. If you have 20 destinations, there are 1,048,576 possible combinations.

D-Wave’s next-generation computer is designed to handle problems with as many as 512 variables. In theory, that lets you solve problems involving two to the 512 possible combinations, and a problem of that size is beyond the reach of any classical computer that could ever be built. “It’s bigger than the number of atoms in the universe,” Rose says. “It doesn’t matter how big a supercomputer you make.”]

With that kind of computing power, Google could go beyond recognizing that you like to go to Starbucks and Best Buy, they could literally predict almost every possible combination of things you could do (commercially) in a given day. Please note the "commercially" part and stop doing naked push-ups in your living room while yelling "try to predict this Google!" You've made your point.

So, is D-Wave's computer quantum, or just a really amazing conventional machine? No one really knows because D-Wave has kept much of the research secret. Scientists are limited to testing it out to see if it displays properties that are consistent with what a quantum machine should be able to do, rather than accessing the data and components directly. They're really whiny about it too, calling it an "enigma, in a negative sense", according to UC math professor Greg Kuperberg and reported by Metz in February. Wambulance.

I'm glad D-Wave didn't fall for the guilt trip, and with orders from Lockheed in May and Google in June, I'd say that it's working out pretty well for them. Quantum or not, the machine seems to work and it's selling. And, if somewhere down the line we all get to use it (or some version of it) as a cloud service, then I don't care, so long as I can ask my computer what I'm doing that day right when I wake up.