A Generic Multiprocessor Task Scheduling Toolkit
|
|
Abstract
A current challenge in the development of parallel applications for distributed
memory platforms is the achievement of a good scalability even for a high number
of processors.
The programming with parallel tasks is a suitable programming technique to
implement such parallel applications consisting of a set of well-defined submodules.
In this programming model, the application can be coded as a parallel program
with mixed parallelism in which the submodules represent parallel tasks each
of which can be executed on one or more processors of a parallel target platform.
The efficient execution of a parallel application is based on a schedule that
considers the structure of the parallel application as well as the computation
and communication performance of the parallel target platform to be utilized
for execution.
The schedule assigns each parallel task a start time and a set of processors
of the parallel target platform for execution.
The goal of this research project is the development of a generic scheduling
toolkit.
The toolkit should provide support to compute schedules for a parallel
application given as a task graph, to compare different scheduling algorithms,
to illustrate the structure of a given task graph, and to analyze the produced
schedules.
This project is supported by the German Research Foundation (DFG).
Project numbers: RU591/9-1 and RU591/9-2.
Publications
-
Kunis, R.: Realisierung einer Schedulingumgebung für gemischt-parallele Anwendungen
und Optimierung von layer-basierten Schedulingalgorithmen, TU Chemnitz, Fakultät für Informatik, Doctoral thesis, 2011. online resource available
|
 |
-
Kunis, R.; Rünger, G.: Optimizing layer-based scheduling algorithms for parallel tasks with dependencies. In: Concurrency and Computation: Practice and Experience, vol. 23, no. 8: pp. 827-849. John Wiley & Sons, Ltd. – ISSN 1532-0634, 2011. DOI: 10.1002/cpe.1674 online resource available
|
|
-
Kunis, R.; Rünger, G.: Task-Block Identification and Movement for Layer-based Scheduling Algorithms. In: Smari, W. (Eds.): Proc. of the 2010 International Conference on High Performance Computing & Simulation (HPCS 2010): pp. 132-139. IEEE – ISBN 978-1-4244-6828-7. Caen, France, 2010.
|
|
-
Kunis, R.; Rünger, G.: Optimization of Layer-based Scheduling Algorithms for Mixed Parallel Applications with Precedence Constraints Using Move-blocks. In: Proc. of the 17th Euromicro International Conference on Parallel, Distributed and Network-based Processing (PDP 2009): pp. 70-77. IEEE Computer Society – ISBN 978-0-7695-3544-9. Weimar, Germany, 2009.
|
|
-
Dümmler, J.; Kunis, R.; Rünger, G.: Layer-Based Scheduling Algorithms for Multiprocessor-Tasks with Precedence Constraints. In: Parallel Computing: Architectures, Algorithms and Applications: Proc. of the International Conference ParCo 2007 (Advances in Parallel Computing, vol. 15): pp. 321-328. IOS Press – ISBN 978-1-58603-796-3. Jülich/Aachen, Germany, 2007. online resource available
|
|
-
Dümmler, J.; Kunis, R.; Rünger, G.: A Scheduling Toolkit for Multiprocessor-Task Programming with Dependencies. In: Proc. of the 13th International Euro-Par Conference (LNCS, vol. 4641): pp. 23-32. Springer – ISBN 978-3-540-74465-8. Rennes, France, 2007. online resource available
|
|
-
Dümmler, J.; Kunis, R.; Rünger, G.: A Comparison of Scheduling Algorithms for Multiprocessortasks with Precedence Constraints. In: Proc. of the 2007 High Performance Computing & Simulation (HPCS'07) Conference: pp. 663-669. ECMS – ISBN 978-0-9553018-27. Prague, Czech Republic, 2007.
|
|
Complete list as BibTeX
Copyright Notice
The documents
distributed by this server have been provided by the
contributing authors as a means to ensure timely dissemination of scholarly
and technical work on a noncommercial basis. Copyright and all rights
therein are maintained by the authors or by other copyright holders,
notwithstanding that they have offered their works here electronically.
It is understood that all persons copying this information will adhere to
the terms and constraints invoked by each author's copyright. These works
may not be reposted without the explicit permission of the copyright holder.
Joint work with
Faculty of Computer Science, Chemnitz University of Technology
Prof. Dr. Gudula Rünger
Raphael Kunis
Jörg Dümmler
