* Quantum computing for heavy data crunching
In computing these days, things always seem to come around to power and heat, and processing efficiency. A Canadian start-up called D-Wave is taking a rather unique approach to the heat and power issue by focusing on quantum computing.
Quantum computing uses quantum mechanics to process data more quickly and accurately. I’m not going to get into the details of quantum physics here, but at the most basic level this is the difference between digital computers and quantum computers: digital computers use bits that can be either a 1 or a 0, while quantum computers use qubits, which can be 0 and 1 at the same time. That flexibility enables quantum computers to solve complex problems that even massive supercomputers can’t solve or can’t solve with 100% accuracy in a reasonable amount of time, D-Wave CEO Herb Martin says.
Last week, D-Wave demonstrated what it calls the world’s first commercially viable quantum computer at the Computer History Museum in Mountain View, Calif. The servers will be available in the middle of 2008, but an online service should be up and running by the beginning of next year, Martin says.
D-Wave executives say their systems will enable organizations to run a wider array of applications, stressing that the quantum computers are not meant to replace traditional servers. The idea is to optimize the capabilities of traditional servers by providing a platform for the processing of very complex data crunching, Martin says.
“The kinds of applications I’m talking about are in operations research, supply chain, anything that’s associated with scheduling – scheduling people, scheduling products, scheduling transportation – applications in quantitative finance and applications in life sciences such as biometrics, protein folding, pattern matching,” he says.
So, for example, if you’re running an application that may need to go to a database to match fingerprints or match face images, the app could hand off that complex process to the quantum computer. In fact, the system that D-Wave is building is a hybrid, meaning that it includes a digital preprocessor where the applications run and interface with the outside world, as well as the quantum processor, which is used as an accelerator to perform the complex operations.
But to get back to the heat and power issue, Martin says that quantum processors would be able to manage the kind of problems now thrown on huge supercomputing clusters with significantly lower power demands.
“We use superconducting electronics, so the system has to be cooled, but it doesn’t take a lot of power compared to digital systems,” he says.
A quantum computer may use 1.5 kilowatts of power, he says. Compare that to a rack of blade servers that can eat up as many as 30 kilowatts of power today.
“The quantum processor really takes almost no power at all; it’s the cooling that takes the power,” he says. “This solution is very green. People with big arrays handling complex problems that are taking thousands of processors have got major problems with recasting applications and then those things consume massive amounts of power and require all kinds of management.”
Quantum computing not only provides the more energy efficient environment, but can also handle traditional apps with almost no modification, Martin says.
It makes the quantum computing idea – if it performs as promised – look mighty interesting indeed. Check out more detail about what D-Wave is planning here.
