Fast fluid simulation using cluster system and graphics processing unit

• Main Author: Lee, Zhuo Qi.
• Other Authors: Henry Johan
• Format: Final Year Project
• Language: English
• Published: 2010
• Subjects: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Computer graphics
• Online Access: http://hdl.handle.net/10356/35721
• Institution: Nanyang Technological University

In recent years, fluid simulation has gained much attention in the computer graphics community. It has been extensively used in graphics and its applications such as movies, games and virtual reality to enhance realism. Several efforts have been made to reduce the computational cost of fluid simulation while retaining visual quality, one of which is the method known as “Lattice Boltzmann Method” (LBM). However, when it is run on a PC, it is still unable to produce simulation result at real time. Thus, an existing CPU based LBM simulation was ported into two platforms: a CPU cluster using Message Passing Interface (MPI), and Graphics Processing Unit (GPU) using Compute Unified Device Architecture (CUDA). The implementation details regarding the two platforms as well as some optimization issues are discussed. In addition, we also discuss and compare the performance of the two systems.