Exploring Parallel Algorithms for Volumetric Mass-Spring-Damper Models in CUDA

Abstract. Since theadvent ofprogrammable graphicsprocessors (GPUs) their computational powers have been utilized for general purpose compu-tation. Initially by “exploiting ” graphics APIs and recently through dedi-cated parallel computation frameworks such as the Compute Unified Device Architecture (CUDA) from Nvidia. This paper investigates multi-ple implementationsofvolumetricMass-Spring-Damper systems inCUDA. The obtained performance is compared to previous implementations uti-lizing the GPU through the OpenGL graphics API.We find that both per-formance and optimization strategies differ widely between the OpenGL andCUDA implementations. Specifically, the previous recommendation of using implicitly connected particles is replaced by a recommendation that supports unstructured meshes and run-time topological changes with an insignificant performance reduction