Dissertation/Thesis Abstract

Developing physics simulations with OpenCL on a graphics processing unit
by Perelygin, Kyrylo, M.S., California State University, Long Beach, 2012, 74; 1522252
Abstract (Summary)

Graphics Processing Units (GPUs) have become increasingly powerful over the last decade. With most of their transistors devoted to processing rather than caching or flow control, GPUs have very high throughput. Programs taking advantage of this architecture can achieve large performance gains. To write programs that can offload the computation onto the GPU and utilize its power, new technologies are needed. Recently, a standard called OpenCL was introduced.

Code written with OpenCL can run on a wide variety of platforms, adapting to the underlying architecture. It is an open, powerful, standard that is easy to learn due to similarities with the C programming language.

Running simulations with OpenCL on the GPU architecture is highly recommended, and based on the experiments conducted, simulating many-particle systems on the GPU can boost performance by factors of ∼2 to ∼75.

Indexing (document details)
Advisor: Lam, Shui-Fung
Commitee: Black, Don V., Ebert, Todd
School: California State University, Long Beach
Department: Computer Engineering and Computer Science
School Location: United States -- California
Source: MAI 51/05M(E), Masters Abstracts International
Source Type: DISSERTATION
Subjects: Computer science
Keywords:
Publication Number: 1522252
ISBN: 9781267977472
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest