Abstract
Climate simulation codes like ECHAM5 consider the dynamics of the Earth system over a long period of time. The huge number of time steps to be propagated restricts the spatial discretization to comparably coarse grid resolutions. This requires scalable parallelizations based on the lowly resolved spatial domain which can cope with the computational effort in the non-parallelizable time domain. The paper on hand gives an overview of the optimization of ECHAM5 on the NEC SX-8 with respect to performance and scalability. Several single CPU and scalability optimization are described. The outcome of the optimizations is demonstrated in detailed performance measurements.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Special section: Climate models at the Max Planck Institute for Meteorology (MPI-M). Journal of Climate 19, 3769–3987 (2006)
Kornblueh, L.: ECHAM5 - an atmospheric climate model and the extension to a coupled model. In: M. Resch, T. Bönisch, S. Tiyyagura, T. Furui, Y. Seo, W. Bez (eds.) High Performance Computing on Vector Systems. Proceedings of the High Performance Computing Center Stuttgart, March 2006, pp. 171–181. Springer, Berlin Heidelberg, Germany (2007)
NEC Corporation: SUPER-UX Performance Tuning Guide. Revision 1.2 (2007)
NEC Corporation: SUPER-UX Shared Memory Parallelization Guide. Revision 1.2 (2007)
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 2009 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Borowski, S. (2009). The Climate Model ECHAM5 on NEC SX-8. In: Resch, M., et al. High Performance Computing on Vector Systems 2008. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-85869-0_15
Download citation
DOI: https://doi.org/10.1007/978-3-540-85869-0_15
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-85868-3
Online ISBN: 978-3-540-85869-0
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)