Improving solutions by embedding larger subproblems in a D-Wave quantum annealer

Area of Science:

• Quantum Computing
• Optimization Algorithms

Background:

• Quantum annealing is a heuristic algorithm for combinatorial optimization problems.
• D-Wave Systems Inc. offers hardware implementations of quantum annealers.
• Limited qubits restrict embedding large problems directly onto quantum annealers.

Purpose of the Study:

• To address the limitations of embedding large problems in quantum annealers.
• To improve the efficiency and accuracy of quantum annealing solvers like qbsolv.
• To propose a novel, fast algorithm for embedding larger subproblems.

Main Methods:

• The study focuses on enhancing the qbsolv method for quantum annealing.
• A new algorithm is proposed to facilitate the embedding of larger subproblems.
• The efficiency of embedding is improved by moving beyond complete graph embeddings for sparse problems.

Main Results:

• The proposed algorithm enables the embedding of larger subproblems than previously possible with qbsolv.
• Embedding larger subproblems leads to more efficient utilization of D-Wave quantum annealer resources.
• Improved embedding strategies result in more accurate solutions for optimization problems.

Conclusions:

• The developed algorithm offers a faster method for embedding larger subproblems in quantum annealers.
• Efficiently embedding larger subproblems enhances the performance of iterative quantum annealing solvers.
• This approach leads to better solutions for large-scale combinatorial optimization problems.