A Tool for Scientific Visualization Based on Particle Tracing Algorithm on Graphics Processing Units

Summary form only given. The need to process large volumes of data associated with the increasing computing power contained in the current graphics cards have encouraged the academic community to produce techniques applied to parallel graphics processing designed for scientific applications. In this context, the goal of this paper is to present an implementation of a scientific visualization tool based on the Particle Tracing Algorithm on graphics processing units using, in particular, the Compute Unified Device Architecture (CUDA). It is also evaluated the performance gains obtained related the combination of some of the settings offered by CUDA.