Hybrid quantum annealing via molecular dynamics

Area of Science:

• Quantum Computing
• Computational Physics
• Optimization Theory

Background:

• Combinatorial optimization problems are computationally intensive.
• Existing classical algorithms face scalability limitations.
• Quantum annealers show promise but require efficient preconditioning.

Purpose of the Study:

• To propose a novel quantum-classical hybrid scheme for efficient large-scale combinatorial optimization.
• To leverage Hamiltonian dynamics of classical flux variables for preconditioning quantum annealers.
• To evaluate the performance and accuracy of the proposed hybrid scheme.

Main Methods:

• Introduction of Hamiltonian dynamics for classical flux variables linked to transverse-field Ising model spins.
• Utilizing molecular dynamics of classical fluxes as a preconditioner.
• Testing the hybrid scheme on MAX-CUT and Ising spin-glass problems.

Main Results:

• The proposed hybrid scheme demonstrates efficient problem-solving capabilities.
• Hamiltonian dynamics effectively preconditioned quantum annealers by identifying critical spins.
• The method showed competitive performance against standard classical algorithms like tabu search and simulated annealing.

Conclusions:

• The novel quantum-classical hybrid scheme offers an efficient approach to large-scale optimization.
• Classical Hamiltonian dynamics serves as a powerful preconditioning tool for quantum annealers.
• This hybridization strategy enhances the practical applicability of quantum computing for optimization tasks.