The December 2022 issue of IEEE Spectrum is here!

Close bar

A First for Quantum Computing?

Canadian company D-Wave has sold its first quantum computer, but doubts remain about exactly how quantum it really is

3 min read
A First for Quantum Computing?

Last week, the Canadian start-up D-Wave Systems announced it has sold its first quantum computer to Lockheed Martin, for a reported $10 million.

It’s exciting to think that after more than a decade of work, quantum computers have emerged from the laboratory and are now poised to shake the world of cryptography and zip through calculations that might take a conventional computer billions of years to perform. That's not the case.

At laboratories around the world, physicists are still struggling to piece together even the most rudimentary quantum computers. Their aim is to build general purpose computers using the quantum equivalent of logic gates. D-Wave has a different goal: it's aiming for more specialized machines that are ideal for tasks like data processing and pattern recognition.

The sale of D-Wave's first system, the D-Wave One, suggests at least one company has confidence in the approach. Lockheed spokesperson Thad Madden told IEEE Spectrum that the firm is hoping to use the system to cut the cost of testing aircraft control systems and other complex combinations of hardware and software that interact with the physical world. “We believe quantum computing can tackle these problems in a more cost-effective and efficient manner than can be done by conventional computers and processes,” Madden says.

Others are not convinced D-Wave has a viable quantum computer, something that will be significantly faster than conventional systems. This was one of the reasons we listed D-Wave as a “Loser” in our January 2010 round-up of technology projects that were set to face significant milestones in the coming year.

The company has begun to address some of the technical concerns about its approach. D-Wave's computers encode information in loops of superconducting wire in which current can flow either clockwise, counterclockwise, or in a quantum superposition of the two. That allows the value of each quantum bit, or qubit, to take on three states, either 0, 1, or both simultaneously.

To perform a computation, programmers encode problems to be solved by tuning the interaction between qubits. Like a marble rolling across a bumpy surface, the system then explores energy configurations. Eventually, the system finds the solution by relaxing into the deepest dip in the landscape, the lowest energy state.

If D-Wave's computers are behaving as designed, qubits should be able to change their state by using quantum tunneling to cut through the energy barrier separating a qubit's "0" and "1" configurations.

One long-standing objection to the system has been that thermodynamics can also accomplish the same task: a qubit may be able to tunnel through the barrier, but if it gets enough heat from the environment, it can simply jump over the hurdle.

But in May, researchers published a paper in Nature that showed D-Wave's qubits can change state below 45 milliKelvin, a temperature that is too low for thermal fluctuations to play a role.

The study showed that D-Wave is technically a quantum computer. But, as Adrian Cho reported in ScienceNOW, there are still doubts that quantum tunneling will be enough to make D-Wave's system faster than classical alternatives. Other quantum behavior, like entanglement between qubits, might be needed to make truly speedy systems, Nature News reported.

"The fundamental problem is that D-Wave has not presented any evidence to the scientific community that the D-Wave One, or any of its other devices, can actually solve any problem faster than a classical computer," MIT's Scott Aaronson, a long-time skeptic of D-Wave's claims, told New Scientist.

D-Wave has been forthcoming about some of the system's powers. This week, D-Wave founder and CTO Geordie Rose (pictured above) told Technology Review that software using the D-Wave systems “can learn things like how to recognize particular objects in photos up to 9 percent more accurately than a conventional alternative”. But speed has yet to be discussed in public. “As a practice, we don’t report performance characteristics of the architecture,” Rose told Forbes earlier this week.

So it may be a while longer before we'll know whether the company could one day make the "Winners" list. More papers from the D-Wave team are reportedly on the way.

(Image: D-Wave Systems)

The Conversation (0)

Why Functional Programming Should Be the Future of Software Development

It’s hard to learn, but your code will produce fewer nasty surprises

11 min read
A plate of spaghetti made from code
Shira Inbar

You’d expectthe longest and most costly phase in the lifecycle of a software product to be the initial development of the system, when all those great features are first imagined and then created. In fact, the hardest part comes later, during the maintenance phase. That’s when programmers pay the price for the shortcuts they took during development.

So why did they take shortcuts? Maybe they didn’t realize that they were cutting any corners. Only when their code was deployed and exercised by a lot of users did its hidden flaws come to light. And maybe the developers were rushed. Time-to-market pressures would almost guarantee that their software will contain more bugs than it would otherwise.

Keep Reading ↓Show less