GPU accelerated grid-free surface tracking.

We present the first GPU accelerated explicit surface tracker. Its speed and the fact that it does not use a grid makes it suitable for a wide range of applications including those that operate on a large, unbounded domain. The core idea for handling topological changes is to detect and delete overlapping triangles as well as triangles that lie inside the volume. The resulting holes are then joined or closed in a robust and efficient manner. We maintain a good quality mesh by using several improvement steps. The algorithm is guaranteed to produce a manifold mesh without boundary. In this paper we describe how all these steps can be parallelized to run efficiently on a GPU. Our results show the quality and efficiency of the method in Eulerian and Lagrangian liquid simulations, in solid simulations and in mesh processing applications. Our GPU implementation runs more than an order of magnitude faster than the CPU version. Graphical abstractDisplay Omitted HighlightsThe first GPU friendly algorithm for explicit surface tracking is proposed.Main contributions: parallel hole-pairing, hole-filling, mesh improvement algorithms.Three case studies: Fluid simulation, solid simulation and mesh processing.Results: The algorithm is between 20 and 50times faster than the CPU version.