Scalable error mitigation for noisy quantum circuits produces competitive expectation values

Abstract

Noise in existing quantum processors only enables an approximation to ideal quantum computation. However, for the computation of expectation values, these approximations can be improved by error mitigation. This has been experimentally demonstrated in small systems but the scaling of these methods to larger circuit volumes remains unknown. Here we demonstrate the utility of zero-noise extrapolation for practically relevant quantum circuits using up to 26 qubits, circuit depths of 120 and 1,080 CNOT gates. We study the scaling of the method for canonical examples of product states and entangling Clifford circuits of increasing size, and extend it to simulating the quench dynamics of two-dimensional Ising spin lattices with varying couplings. These experiments reveal that the accuracy of physically relevant observables after error mitigation substantially exceeds previously expected values. Furthermore, we show that the efficacy of error mitigation is greatly enhanced by additional error suppression techniques and native gate decomposition that reduce the circuit time. By combining these methods, the accuracy of our quantum simulation surpasses the classical approximations obtained from an established tensor network method. These results establish the potential of a useful quantum advantage using noisy, digital quantum processors.