Optimization Algorithms for Quantum and Digital Annealers

Abstract

Quantum-inspired algorithms have the potential to be faster than standard classical algorithms when problem instances exhibit certain structures. In this thesis, we make progress toward improving classical optimization algorithms by borrowing tools from quantum algorithms. We will study implementations that are classical but inspired by quantum devices. Quantum annealers and quantum-inspired optimizers have the potential to provide accelerated computation for certain combinatorial optimization problems. Here, we will examine some quantum-inspired classical implementations and apply benchmarks to the problems in the field of computational biology and material discovery. We will also show that in certain problems, simulations of quantum-inspired algorithms execute computations faster than quantum annealers. We will benchmark the performance of the problems against well-known classical optimization methods for Ising-like models, such as metropolis and greedy algorithms.