Those characterizations hang over a Burnaby company that’s trying to change the world by supercharging the calculating power of computers.
D-Wave Systems says it already has working models of its super-smart machine, a “quantum computer” that relies on the behaviour of subatomic particles of matter, shifting around on a customized chip built by NASA, to work out complex or lengthy problems that conventional computers are not structured to deal with.
Plenty of electronic devices have already been developed on the foundations set out in the century-old science of quantum physics — notably, the television. A computer that relies on the odd behaviour of subatomic quantities of matter is potentially a game changer.
That’s because a quantum-driven computing machine has, in effect, a third gear not available to a regular or “classical” computer.
A classical computer works out problems using on-off switches that either allow or block the flow of electricity. A quantum machine has a third switch option — on-and-off — where the switch is simultaneously open and closed. It’s in the nature of quantum physics to ignore rules that would seem to make no sense at the scale of large objects; for instance, people or planets can only be in one place at a time, whereas subatomic particles can be in two.
“It has been known for 100 years that you could do this, it’s just very hard to engineer,” D-Wave founder and chief technology officer Geordie Rose said in an interview at the company’s office-lab facility.
“In some ways it’s like the introduction of language into human culture.”
Scientists believe quantum computers would be vastly more adept at working out problems or cracking codes and could provide quick answers to questions that it might take billions of years for classical machines to work through.
However, it’s an open question as to whether quantum computers would be superior, since no one other than D-Wave claims to have invented one that works as anything beyond a laboratory experiment.
It is extraordinarily difficult to build a quantum computer to a commercial scale, and the scientific consensus is that it will take another 20 to 50 years for the industry to achieve that milestone. In particular, it is difficult to filter out enough heat and noise to stop the quantum computing process from being disrupted.
Consequently, D-Wave’ claims of success have attracted some heavyweight skeptics, such as the 150,000-member Institute of Electrical and Electronics Engineers, and the Massachusetts Institute of Technology.
If MIT, Los Alamos National Laboratory and IBM can’t do it, how could a British Columbia company with 50-odd employees and a relatively meagre budget?
Lockheed Martin decided to find out: D-Wave announced in May a $10-million multi-year deal with the global security manufacturer to run “verification and validation” tests on its equipment.
“Imagine you build a very complex machine that has many bits and pieces in it, software and hardware — imagine a fighter jet. It could fail in a million different ways, and you can’t test them all,” Rose said.
Classical computing design is reaching a point where long, complex tasks threaten to consume an unsustainable amount of energy and resources relative to the work they perform, Rose said.
“Some of the computers that are on the drawing boards for next-generation high-performance computing systems, so called exo-scale systems, consume an enormous amount of power; literally, a single nuclear power station would be required to power some of these proposals, which is ridiculous.
“So we’re looking for more efficient ways to do that. Quantum computers do provide the possibility of dramatic, revolutionary change in the amount of resources required to do things.”
The machine being tested by Lockheed is contained inside a monolithic black box that could hold an elephant, and resides in the D-Wave office where it was assembled.
“One customer does not a company make, but it’s an important first step,” Rose said.
Key internal components include a cylinder where multiple layers of shielding block noise and magnetic fields, and vacuum and chilling equipment that reduces the temperature around the quantum computing chip to within fractions of a degree of absolute zero — colder than anything you’d find in nature, Rose says.
“Problems that conventional computers can’t solve, quantum computers can’t solve either.
“The thing that people want to know is whether these types of machines can solve the same types of problems much more efficiently, which means with a lot less use of natural resources such as energy, time, memory.
“What we are trying to do is take problems that are very difficult for conventional computers, in the sense that they take a long time or you need a nuclear power station to run them, and execute the same kinds of computation but with dramatically reduced resource needs, either energy or time.”
Cold is the key.
D-Wave has been able to demonstrate, in a paper published in May in the scientific journal Nature, that its system works at temperatures so low that only a quantum effect can be responsible.
Rose said the D-Wave machine is not the only possible design for a quantum computer, but it’s the least-daunting approach to operating one. To date, they’ve obtained 62 patents on their work.
In a commentary accompanying the Nature article, journalist Zeeya Merali reports that one of the skeptics — MIT computer scientist Scott Aaronson — wants more proof.
“There is a lot of dogma in this field that quantum systems are very difficult to build and maintain. It’s absolutely not the case,” said Rose.
“The world is slowly realizing — and I don’t just mean physicists but all people who study nature — that quantum mechanics is not some exotic thing that only particle physics people … have to worry about. In fact it’s everywhere, it’s ubiquitous. It’s not fragile. It doesn’t go away.”
For example, a couple of years ago, researchers found evidence of quantum mechanics in photosynthesis, the process by which plants and bacteria convert sunlight to energy.
“The [Nature] paper was designed to point out, in a very clear way, that a lot of the dogma in this field about how difficult it is to build these things is simply wrong. I believe the paper accomplished that. At least, it’s the beginnings of that narrative.
“We believe that we’ve found a way that’s faster, cheaper, better, to get to the goal of building large-scale quantum computers. Of course there is some resistance¬, there always is, to new ideas. The thing we have on our side is that we are right.”
Source:http://www.vancouversun.com/business/Wave+seeks+engineer+kind+computer/5014156/story.html
