Error-resilient Monte Carlo quantum simulation of imaginary time

Computing the ground-state properties of quantum many-body systems is a promising application of near-term quantum hardware with a potential im-pact in many fields. The conventional algorithm quantum phase estimation uses deep circuits and requires fault-tolerant technologies. Many quantum simulation algorithms developed recently work in an inexact and variational manner to exploit shallow circuits. In this work, we combine quantum Monte Carlo with quantum computing and propose an algorithm for simulating the imaginary-time evolution and solving the ground-state problem. By sampling the real-time evolution operator with a random evolution time according to a modified Cauchy-Lorentz distribution, we can compute the expected value of an observable in imaginary-time evolution. Our algorithm approaches the exact solution given a circuit depth increasing polylogarithmically with the desired accuracy. Compared with quantum phase estimation, the Trotter step number, i.e. the circuit depth, can be thousands of times smaller to achieve the same accuracy in the ground-state energy. We verify the resilience to Trot-terisation errors caused by the finite circuit depth in the numerical simulation of various models. The results show that Monte Carlo quantum simulation is promising even without a fully fault-tolerant quantum computer.

Automated summary

This work proposes an algorithm that combines quantum Monte Carlo with quantum computing to simulate the imaginary-time evolution and solve the ground-state problem. By sampling the real-time evolution operator with a random evolution time according to a modified Cauchy-Lorentz distribution, the expected value of an observable in imaginary-time evolution can be computed. The results show that Monte Carlo quantum simulation is promising even without a fully fault-tolerant quantum computer.