In the quantum computer era we are supposed to see devices that solve complex problems at absurdly fast speeds.
Research and experimentation have gone on for years to develop a quantum computer that leaves ordinary computers in the dust. Unfortunately, new research shows that commercial quantum computers are no faster.
Canadian company D-Wave Systems Inc., based in Burnaby, created the world's first commercial quantum computer and sold it to Google and NASA. While many do believe that the computers produce quantum speedup, or faster performance, others remain skeptical.
Matthias Troyer, a theoretical physicist from ETH Zurich in Switzerland, and the lead author of this study, took a D-Wave Two quantum computer and measured its supposed quantum speedup against a normal computer's processing speed. His findings disappointed those that hoped to find a significantly higher speedup.
Quantum computers process complex problems by performing multiple calculations at the same time, using quantum bits. Quantum bits, or qubits are the particles of quantum devices, just as bits are those used by classical computers. Bits can perform at values of either 0 or 1, but qubits can be 0 and 1, simultaneously.
Quantum computers, like classic computers, solve problems at a lowest energy state. However, the process of finding the lowest energy state is complicated. The exciting feature of quantum computers is that they can find the lowest states faster than ordinary computers, reaching a solution more quickly.
In truth, the D-Wave Two is not a quantum computer, but a quantum annealer, which is only part of a computer. The annealer's role is to specify interactions for its qubits so they can find the lowest energy states. Ordinary computers use thermal annealers whose qubits find the lowest energy states less quickly.
The benefit of the quantum annealer is that it solves specific problems like pattern recognition and machine learning much faster.
Researchers explain that the higher speeds of quantum computers occur when the problems presented are more complex. Thus, the results of any comparisons depend greatly on the kinds of problems researchers choose to give the computers.
Troyer and his team of researchers found that the D-Wave 2 device was no faster than a classical computer, leading some physicists to believe them, and others to think Troyer did not use the right kinds of problems for a quantum speedup. Troyer took the quantum annealer and pitted it against a classical computer's thermal annealer. Ordinarily, researchers would measure how long the devices take to solve problems, and the increases in those times when problem became more difficult. Troyer, however, took a more direct approach and compared the D-Wave's speed of finding its lowest energy state with that of a thermal annealer. He did not find a significantly higher speed in the D-Wave. His findings were published in Thursday's issue of Science.
Troyer and his team remain optimistic. "Our results do not rule out the possibility of speedup for other classes of problems and illustrate the subtle nature of the quantum speedup question", he says in the study.
