GPU simulations of spiking neural P systems on modern web browsers

In this work we present a novel and proof of concept Spiking Neural P system (for short, SN P systems) simulator that runs on modern web browsers whilst using graphics processing units (for short, GPUs). By creating an SN P system that both utilizes the GPU and runs on modern web browsers, we allow a much more performant SN P simulator that would also be a lot more accessible for researchers to experiment with, and can be integrated into other tools or visualizations transparently without having to learn specific GPU knowledge or techniques. Using previous results on representing SN P system computations using linear algebra, we analyze and implement a computation simulation algorithm on web browsers that runs on the GPU. Since web browsers (at this time) do not have any capabilities for General Purpose computing on GPUs (for short, GPGPU), we exploit the Web Graphics Library (for short, WebGL) and create shaders to generate textures that correspond to computational results of our SN P simulation algorithm. To our knowledge, this is the first work on simulating SN P systems on browser GPUs. Here, we present two different implementations and algorithms as case studies to analyse and compare the performance of the simulations, with particular interest in speedup compared to CPU approaches.