high performance computing on graphics processing units: hgpu.org

This paper deals with parallelization of the Quine-McCluskey algorithm. This boolean function minimization algorithm has a limitation when dealing with more than four variables. The problem computed by this algorithm is NP-hard and runtime of the algorithm grows exponentially with the number of variables. The goal is to show that parallel implementation of the Quine-McCluskey algorithm on graphics processing units (GPUs) brings significant acceleration of computing process. Parallelization of the algorithm is implemented through Compute Unified Device Architecture (CUDA), which is a parallel computing platform and programming model created by NVIDIA and implemented by graphics processing units (GPUs).