A blind quantum computer’s laboratory debut

In a blind quantum computation, a client runs an algorithm on a remote server without revealing anything about her input or computation. Now, Philip Walther (University of Vienna) and collaborators have implemented a small-scale version of such a computer. They used a protocol based on so-called one-way quantum computing: Starting with a highly entangled state, the computer performs a series of single-qubit measurements. The measurement results are then processed classically to give the computation output. To make the computation blind, the client herself prepares the qubits with phase angles θ_i that only she knows. She instructs the computer to measure the qubits at angles δ_i. The computer entangles the qubits in a specified way, measures them, and transmits the results back to the client. Without knowing the θ_i, neither the computer nor an eavesdropper can figure out what computation is being done. The Vienna experimenters used photons for their qubits. With just four qubits, pairwise entangled into the linear cluster shown in the figure, the researchers demonstrated various one- and two-qubit quantum logic gates, the eventual building blocks of a larger computation. Importantly, they found blindness doesn’t require all the qubits to be prepared in states unknown to the computer: In their four-qubit system, it suffices to have two qubits (θ₂ and θ₃ in the figure, shown in green and yellow) in secret states. (S. Barz et al., Science 335, 303, 2012.)—Johanna Miller