D-Wave Claims Leap in Quantum Computing, Researchers Skeptical

(Bloomberg) -- D-Wave Systems Inc. has sold its quantum computers to Google and U.S. defense contractor Lockheed Martin Corp. It has now begun selling a new, more-powerful machine.

A single unit of quantum data is called a quantum bit, or "qubit", and the Burnaby, Canada-based company said its new device has a processor capable of handling 2,000 of them at any instant -- twice as many as the previous version. D-Wave also said that initial benchmark tests had shown achievable performance was as much as 1,000 times greater than the predecessor.

"We are seeing an advantage over the best of what you can do classically," Vern Brownell, D-Wave’s chief executive officer, said in a telephone interview, referring to classical computing.

While such an exponential leap in performance is consistent with what researchers have hoped for from quantum computing, D-Wave’s system remains controversial. Some critics say the company’s approach will never achieve the sorts of computational breakthroughs scientists have been expecting from the technology.

Theorists have predicted quantum computing will transform everything from financial portfolio design to genetic engineering to encryption. Their underlying technologies are dramatically different from the kinds of processors found in standard computers, which are approaching the upper limits of their capabilities thanks to demands for them to be smaller but more powerful, faster but less energy-consuming, every year.

D-Wave, founded in 1999, has not said what it will charge for its latest machine, but previous versions are reported to have cost between $10 million and $15 million each.

The company is not alone in building quantum technologies. Google is working on creating a general quantum system at a research center in Santa Barbara, California. Alibaba Group Holding Ltd. said in July last year it had co-founded a laboratory with the Chinese Academy of Sciences to develop quantum computers and security systems. Microsoft Corp. and International Business Machines Corp. also have their own quantum-computing research initiatives.

Limited Benefits

Standard computers process information encoded in a binary format -- represented by either a 0 or 1. This basic building block of processing power is called a bit. By contrast, quantum computers use quantum mechanical principles, such "superimposition" -- meaning its quantum bits can be both 0 and 1 simultaneously. In theory, this should allow quantum computers to solve very difficult algorithms exponentially faster than standard computers.

"You can’t judge a system based on the number of qubits alone," John Morton, a professor of quantum computing at University College London, said. Morton, who had not seen D-Wave’s latest machine but was familiar with earlier versions, said that a major problem with the Canadian company’s approach is that its qubits only remain in a quantum state for about 10 nanoseconds, which is not enough time for the system to solve a very complex equation. As a result, the benefits of such a system over a classical computer is limited, he said.

Another theoretical advantage of quantum computers over classical machines is called "entanglement," or the idea that qubits function collectively in order to achieve huge advantages in cracking hard problems. Morton said entanglement seems to be very limited in D-Wave’s machines. It mattered little, he said, if you had 10 qubits or 2,000 qubits, if only eight were ever entangled.

Most importantly, D-Wave’s system is not a general quantum computer, meaning it can’t solve any type of mathematical problem. Instead, it is useful for solving what are known as optimization problems -- such as planning the most efficient route between a set of points. This is useful in figuring out the most fuel-efficient way to manage an entire fleet of aircraft, for instance. But it cannot perform some of the mathematical wizardry that have caused the greatest hope -- and fear -- about the advent of quantum computing, such as being able to crack most existing forms of public key encryption.

Technical Achievement

Andrew Green, a theoretical physicist at the London Center for Nanotechnology who has run experiments on D-Wave’s previous machines, said that despite some shortcomings, D-Wave has made important breakthroughs. "What they have done is a remarkable technical achievement," he said, comparing it to the Wright Brothers’ first efforts at flight.

Jeremy Hilton, D-Wave’s senior vice president for systems, said that a big advantage of D-Wave’s system over other approaches to quantum computing is that D-Wave’s system does not require lots of error correction, which is not true of other approaches researchers have been trying.

D-Wave has received venture capital funding from Fidelity Investments, Amazon Inc. founder Jeff Bezos’s Bezos Expeditions investment firm, and In-Q-Tel, a venture firm funded by the U.S. Central Intelligence Agency.

To contact the reporter on this story: Jeremy Kahn in London at jkahn21@bloomberg.net. To contact the editors responsible for this story: Nate Lanxon at nlanxon@bloomberg.net, Andrew Pollack