Geek Speak: Vern Brownell, CEO of D-Wave Systems

Vern Brownell says his technology company is developing the world’s first scalable quantum computer. According to its president and chief executive officer, D-Wave Systems will, for the first time, “sell, ship, and install” systems early this year.

D-Wave began working on a superconducting adiabatic quantum computer in 2004. Running such a machine requires a magnetic vacuum and temperatures close to absolute zero (-273 °C). In 2007, the company demonstrated the Orion system, which was criticized by some scientists who questioned whether it truly was a quantum computer. D-Wave reached its latest milestone in December 2009, when Google used one of its C4 Chimera chips in an image-recognition demonstration. According to Brownell, it may have been the “first time a quantum computer was used for any real calculation or any real computation whatsoever”.

Brownell joined D-Wave—which was founded in 1999 and has a staff of about 55 people—in September 2009. The 51-year-old, Massachusetts-born technology executive had moved to Vancouver from outside Boston, where he founded Egenera, a virtualization-technology company, in 2000. From 1989 to 2000, Brownell served as chief technology officer for the New York-based financial-services firm Goldman Sachs.

The Georgia Straight reached Brownell by phone at his office in Burnaby.

What technology is D-Wave working on?

We’re building the world’s first scalable quantum computer. And a quantum computer is basically a computer which uses the laws of nature—or the laws particular of quantum mechanics—to compute. So, we’re harnessing nature really to build a computing device using quantum mechanics.

How is Google using your technology?

Well, Google has really just done a demonstration, where they’ve shown or demonstrated that you can use a technique called machine learning to train a pattern recognition. So, they did a demonstration where they were actually recognizing a car in a picture. And you can use this technology to basically train their algorithms and their computers to do the best job possible of recognizing a car in a picture.

So, you can see there are lots of applications for that kind of technology. But this was the particular demonstration that they did there. And really that’s just one small example of what we call an intractable problem, or a very difficult problem to solve. That is what quantum computers are very good at solving—these very intractable, difficult problems.

What are some of the ways that a quantum computer is different than a regular computer?

Well, a quantum computer uses, rather than bits or binary—0s and 1s—it uses a concept called quantum bits. We call them qubits. And a qubit can be in multiple states at the same time. It’s called superposition. So, it can be in 1 and 0 at the same time, which is a really unique property of quantum mechanics, and something that’s fairly still mysterious to physicists and scientists. So, it’s not completely understood.

But the fact that the computer—this bit—can be in multiple states at the same time leads to some really interesting properties. For instance, you can imagine, if you have many of these bits and you’re solving a problem, in some ways what our computer is doing is taking all the possible states and solving them at the same time and coming up with the correct answer. So, it allows for a level of parallelism that is not possible in what we call classical computers—the computers that we know today.

How will quantum computing eventually affect the kinds of computers people use everyday?

In my view, quantum computers are going to augment what we call the classical computing environment, rather than replace. We think these machines are most useful for these intractable problems—these very, very difficult problems to solve—that sometimes can take years, decades, or even longer for a classical computer to solve. For some of these problems, a quantum computer like D-Wave’s can shave that down by many orders of magnitude, and that’s what’s really exciting about it. So, for these very, very difficult problems, things like protein folding or pattern matching or optimizing scheduling and things like that, which on the surface may not seem like the most difficult problems, they are in fact very computationally intensive. And that’s really what a quantum computer is going to specialize in and really help.

Some of these problems are world changing. You know, when you talk about protein folding and biochemistry and some of the exciting things that can happen, they can have huge positive benefits on mankind based on this kind of computing, which is just not possible with even the best of the high-performance computers and supercomputers that are available in the world today. But it’s not general-purpose computing, and it’s not something that you’re going to have in your laptop or whatever, at least in the next 50 years or so.

How do you respond to skeptics who don’t believe you actually have a quantum computer?

The most important thing that we need to do more of—and we have started aggressively doing this—is basically just working with the scientific community to share our results and to really have them decide whether they think what we’re doing is truly a quantum computer or not. There has been some controversy, and naturally something that has this potential to change the world is going to have skeptics.

It’s only natural. I would be skeptical too. So, it’s incumbent upon us to publish, which we are doing—publishing papers and having those peer-reviewed papers and sharing the results with the scientific community. In fact, our initial customers are going to be some of the most advanced scientific institutions in the world, and we’re well underway with that right now.

How close is your product to commercial release?

We’ll be installing our first machines in the early part of 2010. The market I look at is both scientific and commercial. I think our initial customers are going to be our scientific customers, and then later on we’ll have, you know, broader commercial customers.

This will be the first quantum computer or the first certain kind of quantum computer?

We say the first scalable quantum computer. Because there have been quantum computers that have been built that are like two qubits—I mentioned this concept of a qubit—that really do nothing more than prove that you can build a quantum computer but it didn’t really solve anything or solve a problem.

First of all, what we’ve delivered, even what we demonstrated with Google, is 56 qubits, which is almost 10 times what anybody else has done—maybe more than 10 times. And it’s the first time it’s ever really solved a useful problem. And then the architecture that we’ve built can scale up to hundreds and thousands of qubits, where it really, really starts to get interesting with the ability for this to solve these intractable problems that we talked about. So, that’s why we say scalable.

There have been people that have built in laboratory settings, you know, one- and two-qubit machines using different technologies. But we really do believe that we’re the first to ever build a scalable technology. And certainly at this scale, we can prove that we have nearly 100 qubits working together today.

Every Friday, Geek Speak catches up with someone in Vancouver’s technology sector, video-game industry, or social-media scene. Who should we interview next? Tell Stephen Hui on Twitter at twitter.com/stephenhui.