‘Quantum computer’ would be 100 million times faster than regular computer

Researchers at NASA and Google have been able to solve an optimization problem 100 million times faster with the D-Wave X2 quantum computer than with a conventional computer with a single processor core.

With the research, the NASA and Google employees want to show that a quantum annealing computer such as the D-Wave can actually be more efficient than a classical computer. The study compared the performance of the D-Wave quantum computer with that of a classical computer using so-called simulated annealing. The quantum approach turned out to be orders of magnitude faster. The simulation was done with an algorithm called Quantum Monte Carlo.

There are, however, a few caveats to be made about the results. The most important is that the quantum mechanical functioning of the D-Wave quantum computer is still not conclusively proven. Moreover, the machine can only be regarded as a quantum computer to a limited extent, since only quantum annealing is possible with the qubits. In addition, a quantum annealing computer is only suitable for solving a limited series of problems, such as the traveling salesman problem. Translating problems into an issue compatible with the D-Wave is also not easy. The result obtained would be encouraging for the feasibility of a quantum mechanical approach to complex problems.

It is not the case that practical applications are in the offing, Google writes on its research blog. With the study, which appeared on the arXiv site and has not yet been subjected to the judgment of other scientists, the team aims to demonstrate that a problem involving nearly a thousand binary variables can be solved 108 times faster than a normal processor core using simulated annealing. MIT Technology Review notes that the conventional computer running the code had to use a similar algorithm to that used by the D-Wave. However, an alternative algorithm is known for conventional computers that would make them much faster and perhaps even faster than the D-Wave for this particular problem, according to MIT Technology Review.

The D-Wave X2, the third generation of computers from Canadian D-Wave Systems, was used for the research, with which certain quantum algorithms can be performed via quantum annealing. The heart of the computer operates at a temperature of 15 millikelvin, or just above absolute zero of -273.15 °C.

Time it took to find an optimal solution with 99 percent probability for different problem sizes. Source: Google Research