Verifying the accuracy of quantum computers

Science : To solve certain problems that are too laborious for classical computers, researchers have been developing quantum computers. Such computers rely on particles’ quantum mechanical phenomena, such as superposition and entanglement, to perform calculations. Because a unit of quantum information—a qubit—can embody many different states simultaneously, quantum computers can run through every possible combination of qubits instantaneously and thus provide superfast processing times. Current quantum computers use just a few qubits, and their calculations can be checked by classical computers. However, as quantum computers become larger and more complex, verifying their accuracy could pose problems. To resolve that issue, Stefanie Barz of the University of Vienna and colleagues propose a technique called blind quantum computing, which involves setting a series of “traps,” or short, intermediate calculations whose results are already known to the user. As long as the computer comes up with the correct results for those calculations, the user will know it is working properly. By using blind quantum computing, any quantum computer, regardless of size, should be able to double-check any other.