Experts doubt quantum computer's authenticity

Posted by Humphrey Cheung

Mountain View (CA) - D-Wave's quantum computer demonstration this week has generated lots of buzz, but some experts doubt the authenticity of the demo. At the Computer History Museum in Mountain View California, company reps solved Sudoku problems and calculated seating plans, but some university researchers and corporate scientists need to see more data to be convinced. They are also concerned about D-Wave's lack of peer research and of the test's remote nature. 

D-Wave demoed the computer on Tuesday and solved several complex problems that would have required a good amount of traditional computer power. Orion, the company's code-name for the prototype computer, appeared to solve a Sudoku puzzle, along with calculating a complicated seating plan which had several rules like Guest A couldn't sit near Guest B, but Guest B had to sit with C. Orion supposedly has 16-qubits, or quantum binary bits of power and D-Wave says that could reach 1024-qubits in two years.

Unlike traditional bits that represent just 1 and 0, quantum bits could represent an almost infinite number of variations, making computers built on quantum technology good for solving cryptography, pattern matching and database search problems.

While impressive, some scientists don't like the fact that the calculations were actually being done at a remote location and that the computer couldn't be physically inspected. D-Wave reps were actually quite open about that detail saying the computer was too delicate to be moved because of its liquid helium cooling system and sensitive components.

Lieven Vandersypen, an associate professor at Delft University and quantum computing researcher, told the IEEE Spectrum that he is surprised that investors have put money into the company and that D-Wave "hasn't published any major advances or breakthroughs in the scientific literature." He adds that the company has very little detail to support their claims, something that a peer-reviewed article would have.

Phil Kuekes, a computer architect in the Quantum Science Research Group at HP Labs, was also skeptical telling the Associated Press, "Until we see more actual measurements, it's hard to know whether they succeeded or not."

According to the Associated Press, even D-Wave's own Chief Executive Herb Martin says the machine isn't a real quantum computer, but is instead a "kind of special-purpose machine that uses some quantum mechanics". But despite this admission, D-Wave still expects a commercial quantum computer will be available in 2008. Like traditional mainframes, computation time would be rented out.