Algorithmic simulation of far-from-equilibrium dynamics using quantum computer

TL;DR

This paper demonstrates the potential of superconducting quantum computers to simulate far-from-equilibrium spin dynamics, revealing their ability to reproduce key physical phenomena despite current limitations.

Contribution

It provides the first proof-of-principle digital simulations of spin models on superconducting quantum computers, highlighting their capabilities and challenges.

Findings

Quantum computers can simulate spin dynamics and reproduce symmetry effects.

Current hardware limitations affect the length and accuracy of simulations.

Heuristic methods can extract useful information from imperfect data.

Abstract

We point out that superconducting quantum computers are prospective for the simulation of the dynamics of spin models far from equilibrium, including nonadiabatic phenomena and quenches. The important advantage of these machines is that they are programmable, so that different spin models can be simulated in the same chip, as well as various initial states can be encoded into it in a controllable way. This opens an opportunity to use superconducting quantum computers in studies of fundamental problems of statistical physics such as the absence or presence of thermalization in the free evolution of a closed quantum system depending on the choice of the initial state as well as on the integrability of the model. In the present paper, we performed proof-of-principle digital simulations of two spin models, which are the central spin model and the transverse-field Ising model, using 5- and 16-qubit superconducting quantum computers of the IBM Quantum Experience. We found that these devices are able to reproduce some important consequences of the symmetry of the initial state for the system's subsequent dynamics, such as the excitation blockade. However, lengths of algorithms are currently limited due to quantum gate errors. We also discuss some heuristic methods which can be used to extract valuable information from the imperfect experimental data.