high performance computing on graphics processing units: hgpu.org

The Boolean Satisfability Problem is one of the most important problems in computer science with applications spanning many areas of research. Despite this importance and the extensive study and improvements that have been made, no efficient solution to the problem has been found to the date. During the last years, nVidia introduced CUDA, a platform which lets developers take advantage of the great computing power of the GPUs, that was once unavailable to use. In this work, we propose a methodology of using GPUs to accelerate the resolution of the SAT problem. Since the problem does not map well the GPU paradigm, we will explain our approach, which uses both the GPU and the CPU to solve the parts that suit them the best. We also propose developing a novel SAT solver, capable of solving current benchmark problems, as well as most industry problems.