The CEO of a quantum computing company walked onto a stage at MIT and stood in front of an audience of professors, engineers and computer scientists.
Vern Brownell, CEO of D-Wave Systems Inc., looked out at the crowd and said, "I cannot explain how quantum computing works."
Was he heckled? Did attendees get up and leave? No.
No one in the audience stirred. There was no murmuring. Nobody laughed. No sidelong glances. Nothing.
Quantum computing is just that confusing. Some of the world's best physicists don't understand how it works.
Nobel Laureate physicist Richard Feynman is widely quoted as saying, "If you think you understand quantum mechanics, you don't understand quantum mechanics."
Despite how complex the idea of a quantum computer is and the fact that some physicists say we're as much as 50 years away from seeing one, D-Wave, a quantum computing company based in Burnaby, British Columbia, said it is building them.
NASA, Google and Lockheed Martin are testing them.
"If you want to buy a quantum computer, I can sell you one today," Brownell said at an MIT Tech Conference in February that focused on disruptive technologies.
Those kinds of statements have created a buzz in both the world of physics and the world of computing because many believe quantum computing holds a lot of promise. It's like the Holy Grail of supercomputing.
Think of a computer that could surpass the top classic supercomputers in some calculations, especially problems where you have to search through a lot of data, finding answers to questions so complex that machines like IBM's Blue Gene and Cray's systems might need hundreds of years to solve, or might never solve them.
D-Wave CEO Vern Brownell
Quantum computers might help researchers seeking cures for cancer, advance cryptography or find distant planets. They also could be used to simulate political and military situations, such as the unrest in Ukraine, enabling researchers or a government to test different options and see how they would affect the outcome.
Quantum computers rewrite the rules of how computing works.
Classic computers use bits -- ones and zeroes -- for processing instructions, and they work based on a series of instructions. Ask the computer a question, and it will move through the calculation in a linear, orderly way.
A quantum computer combines computing with quantum mechanics, one of the most mysterious and complex branches of physics. The field was created to explain physical phenomena, like the odd actions of subatomic particles, that classical physics fails to do.
One of the rules of quantum mechanics is that a quantum system can be in more than one state at once. But that concept goes against what's known of the world. Something can be green or red but it cannot be green and red at the same time.
That's not the case with quantum mechanics.
Each bit in a quantum machine -- known as qubits -- can be both a one and zero. It's about possibilities. When a qubit is constructed, it's built so you don't know if it's a one or a zero. It has the possibility of being both.
The D-Wave system with the 512 qubit chip is being tested by NASA and Google. (Photo: D-Wave)
It's not known what those qubits are until they begin to interact - or entangle - with other qubits. Based on their entanglements, they become a one or a zero. However, just because a qubit acted as a zero during one calculation, doesn't mean it will act as a zero during the next calculation. It goes back to the original possibility.
WPI Prof. Germano Iannacchione
That's where the quantum computer's power comes into play.
A quantum system doesn't work in an orderly, linear way. Instead, its qubits communicate with each other, through entanglement, and they calculate all the possibilities at the same time.
That means if a quantum machine has 512 qubits, it's calculating at 2 to the 512th power at the same time. That number is so immense that there are not that many atoms in the universe, according to Rupak Biswas, chief of NASA's Advanced Supercomputing Division. Some physicists theorize that all those calculations are being done in different dimensions.
"We're so far outside of our everyday experience," said Germano S. Iannacchione, head of the Physics Department at Worcester Polytechnic Institute. "Common sense doesn't guide us here. We're trying to come up with pictures in our heads of how it works. When you're at the hairy edge of the unknown in physics and you don't have experience and common sense to guide you, you have to rely on the math. That's the only thing you can hold on to."
Despite the complexities, D-Wave's Brownell said his company has built quantum computers, using their own quantum processor built with different metals, such as niobium, a soft metal that becomes superconducting when cooled to very low temperatures.
One machine, the D-Wave Two, leased by the Universities Space Research Association, is based at NASA's Ames Research Center in Mountain View, Calif. NASA has use of the machine 40% of the time, Google has another 40% and the research association has 20%.
Google declined to talk about its work with the system. However, its experiments on the computer have led to debate on whether D-Wave's computer performs any better than classic computing or whether it is a quantum computer at all.
NASA, which has had its hands on the D-Wave Two since last September, has only been testing it, Biswas said. His group has been doing high-performance modeling and simulation on problems related to Earth sciences, aeronautics and deep space exploration.
"We're still in the early stages," said Biswas, but added that testing is going well. "We are trying to see what it can do. It's not a turnkey situation. It's a very exotic field. It's like in the early days of computing when we had computers with vacuum tubes and card readers."
D-Wave's system at NASA may be the first commercially available quantum computer, but it's not the first quantum machine. Basic quantum computers have been built before.
In 2000, scientists at the Los Alamos National Laboratory demonstrated a working 7-qubit system.
In 2011, Brownell said, D-Wave, to prove it was on the right track, built a quantum computer running 8 qubits. However, the company hasn't proved that its 512-qubit machine works as a quantum computer, and that's because, he said, it simply can't be proven.
"These are such complex systems they can't be modeled by all the computers in the world put together," Brownell said. "That will never be completely provable."
Paul Benioff, who is credited with being the first to apply the theories of quantum mechanics to computers in the early 1980s while working at the Argonne National Laboratory, is doubtful that D-Wave has built a true quantum computer.
"We're a long ways away," he told Computerworld. "It won't happen in my lifetime and I don't intend to die tomorrow."
Benioff says it could be 20 to 50 years before anyone is able to get a lot of qubits to work together.
"It's not hard to build [a qubit], but how do you build a whole lot of them and have one over here interact with one way over there?" he asked. "There are a lot of questions out there about whether they are full quantum computers. It could be a step there or it's an offshoot of the right way to go."
Iannacchione agrees that D-Wave's system is likely a step toward building a real quantum computer.
"They haven't demonstrated the ability to do these huge calculations," he said. "There's no clear evidence that what D-Wave is doing is faster than what a classical computer can do. If they really are creating a quantum computer, it should be hugely faster even if we don't understand what is going on under the hood."
That is what is making many people, in both physics and computer science, skeptical about D-Wave's machine. This is so new and out-of-the-box, that they're not even sure if a true quantum computer has been built.
And having a quantum computer won't be as easy as adding more racks to a company's data center.
A quantum computer has to be completely isolated from everything from radiation to light, heat and even vibrations. It also has to operate at 458 degrees below zero Fahrenheit.
"It's the world's most delicate souffl," said Iannacchione.
If it's difficult to find software to take advantage of computers running multi-core chips, finding software to run on a quantum machine would be a much bigger issue.
Despite the doubts and difficulties associated with quantum computing, Brownell maintains that D-Wave is the first to build a commercial quantum computer that can do large, useful calculations.
"The most complex thing ever done by a quantum computer before ours was factoring the number 21 in a laboratory," he said. "This is one of the most important things to happen in computer science in the last 50 years. This becomes a whole new branch of computer science."
Sharon Gaudin covers the Internet and Web 2.0, emerging technologies, and desktop and laptop chips for Computerworld. Follow Sharon on Twitter at @sgaudin, on Google+ or subscribe to Sharon's RSS feed. Her email address is [email protected].
Read more about high performance computing in Computerworld's High Performance Computing Topic Center.