Distributed Quantum Computing Architectures

Despite dramatic improvements in quantum gate fidelities, coherence times and qubit counts, the capabilities of quantum processors will remain modest in the near future. We compare two approaches that use small unreliable quantum processors to solve large computational problems. The first approach uses hybrid quantum-classical architectures where larger quantum circuits are broken into smaller sub-circuits that are evaluated separately, either using small quantum processors or a quantum simulator running on a classical supercomputer. The second approach leverages quantum networks to teleport shared qubits between quantum processors. We compare the suitability of these techniques for various quantum circuits, describe optimizations that enable mapping quantum circuits into sub-circuits, and compare the hardware requirements of the two approaches.