Quantum simulation of mixed-field Ising model dynamics using pulse-level-controlled Trotter circuits and zero-noise extrapolation

Quantum computers promise to enable the efficient simulation of quantum many-body systems with resources that do not increase exponentially with system size. However, uncorrected quantum noise in current noisy intermediate-scale quantum (NISQ) hardware results in the accumulation of substantial errors during a quantum computation, severely limiting their use. Here, we address this issue using a combination of error mitigation strategies on IBM QPUs and discuss the scaling of their performance with the number and quality of the qubits. We simulate the dynamics of the nonintegrable mixed-field Ising model in a regime where the system exhibits persistent oscillations for certain initial states. These coherent oscillations result from a phenomenon known as quantum many-body scars and provide a useful benchmark for the accuracy of the simulations. By exploiting pulse-level control and implementing several error mitigation techniques, including zero-noise extrapolation, dynamical decoupling, Pauli twirling, and symmetry-based postselection, we are able to follow the many-body coherent oscillations over a longer period of time.