Gesellschaft für Informatik e.V.

Lecture Notes in Informatics


9th Workshop on Parallel Systems and Algorithms (PASA), Workshop of the GI/ITG Special Interest Groups PARS and PARVA, held at the 21st Conference on the Architecture of Computing Systems (ARCS), February 26th, 2008 in Dresden, Germany P-124, 83-92 (2008).

Gesellschaft fuer Informatik, Bonn
2008


Editors

Wolfgang E. Nagel (ed.), Rolf Hoffmann (ed.), Andreas Koch (ed.)


Copyright © Gesellschaft fuer Informatik, Bonn

Contents

Parallel derivative computation using ADOL-C

Andreas Kowarz and Andrea Walther

Abstract


Derivative computation using Automatic Differentiation (AD) is often considered to operate purely serial. Performing the differentiation task in parallel may require the applied AD-tool to extract parallelization information from the user function, transform it, and apply this new strategy in the differentiation process. Furthermore, when using the reverse mode of AD, it must be ensured that no data races are introduced due to the reversed data access scheme. Considering an operator overloading based AD-tool, an additional challenge is to be met: Parallelization statements are typically not recognized. In this paper, we present and discuss the parallelization approach that we have integrated into ADOL-C, an operator overloading based AD-tool for the differentiation of C/C++ programs. The advantages of the approach are clarified by means of the parallel differentiation of a function that handles the time evolution of a 1D-quantum plasma.


Full Text: PDF

Gesellschaft fuer Informatik, Bonn
ISBN 978-3-88579-218-5


Last changed 04.10.2013 18:16:41