Foreword
Recently, our team submitted a research paper called “Accelerating Simulated Quantum Annealing with GPU and Tensor Cores,” which was accepted by research paper of ISC high performance, 2022.
In this paper, we modify the SQA algorithm to increase the parallelism of the SQA algorithm and still guarantee the performance of the accuracy.
In general, SQA algorithms are often used to solve combinatorial optimization problems. We will first map the combinatorial optimization problem (e.g. traveling salesman problem (TSP)) that needs to solve to the appropriate Ising model. Different problems need to be mapped to different Ising models. Some Ising models like the fully-connected model, King graph model, EA model, etc., are referred to by the teams solving quantum annealing-related works (e.g. Tohoku’s SQA accelerated by FPGA and GPU, Toshiba Simulated Bifurcation Machine SBM).