Back to results

University of Mississippi

Graphics Processing Unit Acceleration Of Computational Electromagnetic Methods

Abstract

dc:description.abstract

The use of Graphical Processing Units (GPU's) for scientific applications has been evolving and expanding for the decade. GPU's provide an alternative to the CPU in the creation and execution of the numerical codes that are often relied upon in to perform simulations in computational electromagnetics. While originally designed purely to display graphics on the users monitor, GPU's today are essentially powerful floating point co-processors that can be programmed not only to render complex graphics, but also perform the complex mathematical calculations often encountered in scientific computing. Currently the GPU's being produced often contain hundreds of separate cores able to access large amounts of high-speed dedicated memory. By utilizing the power offered by such a specialized processor, it is possible to drastically speed up the calculations required in computational electromagnetics. This increase in speed allows for the use of GPU based simulations in a variety of situations that the computational time has heretofore been a limiting factor in, such as in educational courses. Many situations in teaching electromagnetics often rely upon simple examples of problems due to the simulation times needed to analyze more complex problems. The use of GPU based simulations will be shown to allow demonstrations of more advanced problems than previously alloby adapting the methods for use on the GPU. Modules will be developed for a wide variety of teaching situations utilizing the speed of the GPU to demonstrate various techniques and ideas previously unrealizable.

Degree

thesis:*
Name thesis:degree_name
Ph.D. in Engineering Science
Level thesis:degree_level
Dissertation
Discipline thesis:degree_discipline
Electrical Engineering
Year dc:date.available
2013

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Inman, Matthew Joseph
Contributors dc:contributor
  • Atef Z. Elsherbeni
  • William Staton
  • Ramanarayanan Viswanathan

Subjects

dc:subject × 4

Identifiers

dc:identifier.*
Repository record dc:identifier
https://egrove.olemiss.edu/etd/528
OAI identifier oai:identifier
oai:egrove.olemiss.edu:etd-1527

Chain of custody

source
Harvested from
University of Mississippi
Base URL
egrove.olemiss.edu/do/oai/
Last updated
2026-07-24
Source record
OAI-PMH GetRecord
citation

Inman, Matthew Joseph. Graphics Processing Unit Acceleration Of Computational Electromagnetic Methods. Dissertation thesis, 2013. https://egrove.olemiss.edu/etd/528