The coldest place in the universe can be found right here in Burnaby.
The world's leader in developing computers that successfully work on quantum mechanical theory, often referred to as the magic of science, is on Still Creek Drive with more than 90 U.S. patents to its name.
D-Wave Systems Inc. sold its first computer to Lockheed Martin and in a recent round of investments, got the backing of In-Q-Tel, which is a venture-capital company that represents the CIA.
Recently, Lockheed Martin announced that it intends to upgrade its D-Wave computer to a commercial scale.
The company was founded in 1999 by UBC scientists and has worked its way up to its current 512 quantum-bit (qubit) processor.
Each system has a price tag of $10 million, is hundreds of thousands of times faster than a conventional computer and requires a temperature colder than deep space to work.
What sets a quantum computer apart from a regular one, besides the fact it's based on complex theories that occur on an infinitesimal level, is its ability to solve optimization problems.
Jeremy Hilton, D-Wave vice-president of processor development, said optimization problems are useful for many things, such as machine learning, cancer detection, radiographs and image-labelling.
"This technology is very relevant to a lot of the major challenges that humanity is facing right now and will be facing moving forward as technology gets more complex," Hilton told the Burnaby NOW. "We're trying to solve ever more complex problems. We believe this is going to be a critical piece of technology that enables us and humanity to be able to solve these kinds of problems."
D-Wave recently worked with Google on a classifier that was able to identify specific objects in an image, known as image-labelling.
"(Computers) think of an image as a set of pixels and ... we're talking about objects which are aggregates of pixels - conveying to a machine how to do that is a very, very difficult thing," Hilton said.
Google offered up its servers and with D-Wave's quantum computer, the classifier was able to pick up specifically on a car in the image.
"It was actually quite successful," Hilton said. "The classifier was slightly better than the normal algorithms that they had used at the time."
It's something that conventional computers would never be able to do, according to Hilton.
"That's a very hard thing, and machines are terrible at it," he said. "They don't work the way our brains do.
"Your brain works in this incredible way to process objects in things that you perceive and it's how we're wired, but computers are terrible at that."
Image-labelling is part of a broader class of machine learning, where machines are able to do what humans can.
"And that's not intelligence per say, but there are elements of processing that's just different from what's possible right now," Hilton said.
The company developed all of its technology from the ground up - right here in Burnaby.
"We've come a long way," Hilton said.
Two years ago, D-Wave came out with its 128-qubit processor and now has its 512-qubit processor, and Hilton says that the way these optimization problems scale is exponential.
"So this processor is four times the number of quantum bits, but the problems are significantly more challenging than they were at the 128-qubit scale," he said. "We expect later this year to see some performance information come out about the 512-qubit processor that we think will be pretty exciting."
For a quantum bit processor to work it cannot be interrupted by noise, which means the temperature has to be extremely low - colder than deep space.
"The D-Wave technology ... sits at almost absolute zero, 10 miliKelvin or minus 273 degrees Celsius, and sits inside of a large (dilution) fridge that allows it to achieve those temperatures," Hilton said.
A dilution fridge is a cyrogenic device normally only found in well-funded physicist laboratories - but in Burnaby, D-Wave has several in its lab made by their own staff.
Despite the processors getting faster and faster, it uses the same amount of power through the fridge, which is a green aspect of quantum computing, according to Hilton.
"When the fridge is running and the processor is running, information is loaded onto the local processor," he said.
This is when quantum mechanics comes into play.
"The quantum bits use a combination of super position, tunneling and things like entanglement to process information in this unique way and solve these optimization problems," Hilton said.
Conventional computers can only have a 0 or a 1 at once in its binary sequence, but a quantum computer can hold a 0 and a 1 at the same time.
Although it doesn't sound like much, it's basically done the unthinkable when it comes to computer technology.
"In this case, the problem is solved a series of 512 bits and will be read out of the processor and will feed back into an algorithm maybe using this information as part of solving a larger problem," he said.
However, D-Wave has been criticized for not using pure quantum mechanics by physicists, and for having computers that are both chunky and expensive.
Visually, a quantum computer resembles more of an old-time computer out of a B-rated science fiction movie - when computers were the size of rooms.
"The computers are obviously very big, they're very expensive," he said. "The challenge is that in the early days, you have a significant amount of work going into producing those technologies.
"We've had to develop all of these technologies."
Hilton said D-Wave computers have proven to successfully utilize quantum mechanical theories that are integrated into the design of the processor, which can barely be seen by the naked eye.
The kind of quantum computing done at D-Wave works with nature and not against it, he added.
D-Wave is currently waiting for peer reviews of its 512-qubit processor. It is also in the process of moving to a larger facility on Beta Avenue., as no other place in the Lower Mainland has space, such as in Burnaby, that fits the needs of quantum computing so well, according to Hilton.
When asked if quantum computers will take the same path as conventional ones, by once being room-size to eventually being hand-sized, he said nothing's impossible.
"I can't say that it's not possible because too many people back at that time ... said it wasn't possible - and they were wrong," Hilton said. "I think where quantum computing will touch the everyday user is in things like the classifier ... (it's) useful to the everyday person, and those will be running on your phone."
For more information, visit www.dwavesys.com.
© Copyright 2013