Fraunhofer-Gesellschaft

Publica

Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Efficient AMG on heterogeneous systems

 
: Kraus, Jiri; Förster, Malte

:
Preprint urn:nbn:de:0011-n-1945705 (202 KByte PDF)
MD5 Fingerprint: d233d0f2b9fc98e823cdcd21020ad254
Created on: 24.2.2012


Keller, R.:
Facing the multicore-challenge II : Aspects of new paradigms and technologies in parallel computing; Second Conference for Young Scientists - "Facing the Multicore Challenge II", held at the Karlsruhe Institute of Technology (KIT), September 28-30, 2011
Berlin: Springer, 2012 (Lecture Notes in Computer Science 7174)
ISBN: 3-642-30396-X
ISBN: 978-3-642-30396-8
ISBN: 978-3-642-30397-5
ISSN: 0302-9743
pp.133-146
Conference "Facing the Multicore-Challenge" <2, 2011, Karlsruhe>
English
Conference Paper, Electronic Publication
Fraunhofer SCAI ()
LAMA; AMG; GPU; CUDA

Abstract
In many numerical simulation codes the backbone of the application covers the solution of linear systems of equations. Often, being created via a discretization of dierential equations, the corresponding matrices are very sparse. One popular way to solve these sparse linear systems are multigrid methods - in particular AMG - because of their numerical scalability. As the memory bandwidth is usually the bottleneck of linear solvers for sparse systems they especially benefit from high throughput architectures like GPUs. We will show that this is true even for a rather complex hierarchical method like AMG. The presented benchmarks are all based on the new open source library LAMA and compare the run times on different GPUs to those of an efficient OpenMP parallel CPU implementati on. As the memory access pattern is especially crucial for GPUs we have a focus on the performance of different sparse matrix formats.

: http://publica.fraunhofer.de/documents/N-194570.html