| CPC G06N 10/60 (2022.01) [G06F 9/4403 (2013.01); G06N 10/40 (2022.01)] | 9 Claims |
|
1. A method of performing computation in a hybrid quantum-classical computing system comprising a classical computer and a quantum processor, comprising:
selecting a maximum-cut (MaxCut) problem to be solved;
executing iterations, each iteration comprising:
setting a quantum processor in an initial state, wherein the quantum processor comprises a plurality of trapped ions, each having two frequency-separated states defining a qubit;
computing a set of first variational parameters and second variational parameters, wherein the set of first variational parameters and second variational parameters is selected randomly in the initial iteration, and updated from the set of first variational parameters and second variational parameters in the previous iteration using a chaotic map that comprises a tuning parameter, wherein the tuning parameter is adjusted to maximize an approximation ratio of an approximate solution relative to the exact solution of the MaxCut problem;
computing a plurality of model Hamiltonian circuits using the selected first variational parameters, and a plurality of mixing circuits using the selected second variational parameters, wherein each of the plurality of model Hamiltonian circuits implements a model Hamiltonian to which the MaxCut problem corresponds;
applying a trial state preparation circuit to the quantum processor to transform the quantum processor from the initial state to a trial state, wherein the trial state preparation circuit comprises the plurality of model Hamiltonian circuits and the plurality of mixing; and
measuring an expectation value of the model Hamiltonian on the quantum processor; and
outputting, by a classical computer, the measured expectation value of the model Hamiltonian as an optimized solution to the MaxCut problem.
|