Year of Award


Document Type

Thesis - Campus Access Only

Degree Type

Master of Science (MS)

Degree Name

Computer Science

Department or School/College

Department of Computer Science

Committee Chair

Jesse Johnson

Commitee Members

George McRae, Joel Henry


cuda, gpu, sph


University of Montana


Recent advances in graphics processing units (GPUs) have exposed the GPU as an at- tractive platform for inherently parallel programming problems. NVIDIA's Compute Unifed Device Architecture (CUDA) has enabled developers to harness the massive computational power of the GPU through the CUDA parallel programming model. Smoothed particle hydrodynamics (SPH) simulations are particularly well suited to CUDA implementation due to the high number of data-parallel computations per- formed. Today's GPUs are powerful enough to power SPH simulations at interactive rates and have made possible the emerging field of real-time uid dynamics. Using the SPH method I have implemented my own real-time SPH simulation in C++ us- ing OpenGL and then integrated CUDA into my existing simulation via the OpenCL framework. This paper focuses on the implementation of a serial SPH simulation running on the CPU and the transition to an implementation running on the GPU.

This record is only available
to users affiliated with
the University of Montana.

Request Access



© Copyright 2011 steven santos garcia