Title:
Implementing the Multi-level Fast Multipole Algorithm on Graphics Processors
Date:
Friday, October 8, 2010 @ 12:30pm
Location:
E2-393, EITC Building, University of Manitoba Fort Garry Campus
Speaker:
Jonatan Aronsson
Abstract:
The Multi-level Fast Multipole Algorithm (MLFMA) is established as the most important method for solving large-scale electromagnetic scattering problems. It has been shown that it reduces the complexity of matrix-vector multiplication to O(NlogN) when using method-of-moments to solve scattering problems. In order to solve larger and more complex problems, research into new ways of parallelizing the MLFMA and improving its performance is very active. In this presentation, we discuss the parallelization and implementation of the MLFMA on commodity graphics processing units (GPUs). The floating-point performance of GPUs is an order of magnitude higher than general purpose processors (CPUs) which makes GPUs an attractive platform for running the MLFMA.
Speaker Bio:
Jonatan Aronsson was born in Olofstrom, Sweden. He received the M.S. degree in engineering physics in 2003 from the Lund Institute of Technology, Lund, Sweden. He spent the 2002-2003 academic year as a visiting student at the University of Waterloo, Waterloo, ON, Canada. He is currently pursuing the Ph.D. degree in electrical and computer engineering at the University of Manitoba, Winnipeg, MB, Canada.
From 2004 to 2006, he held a position as a Systems Manager and Programmer at the University of Waterloo. Since 2006, he has been with the University of Manitoba, Winnipeg, MB, Canada, where he is a HPC Applications Analyst. His current research interest is in fast and parallel algorithms for computational electromagnetics.
Cost:
This will be a free event.
Contact:
For questions or more information: Vladimir Okhmatovski 480-1432
