A Survey of Parallel Simulation of P Systems with GPUs

Abstract

P system simulators become essential for model verification and validation, since they reproduce the semantics of the models in an automatic way. For this reason, in the literature, many authors have proposed several simulation tools. However, in order to handle large instances in an efficient way, parallel simulators come into play. High Performance Computing is a research branch that brings efficient tools for scien- tific purposes. For decades, many parallel platforms and architectures have been designed, with the goal of accelerating compute-demanding applications. But it was 10 years ago, that this field was revolutionized with the dawn of GPU computing through CUDA. This technology allowed programmers to run general-purpose parallel code in GPUs, harnessing in a simplified manner the large amount of processors within a GPU. Many authors have chosen this technology for accelerating the simulation of their P system models. Recently, this topic has captured the attention of more researchers. Therefore, in this paper we survey the related work on GPU-based simulators for P systems, and its evolution over the time until today