Dynamic Component Extension: a Strategy for Performance Improvement in Multicomponent Applications

Multicomponent application paradigms have gained prominence in many significant multidisciplinary scientific applications. In this work, we propose a software strategy called dynamic component extension for multicomponent applications to improve application performance by minimizing processor idling. In this strategy, the processor space of a component is dynamically extended to include the processors of other components during certain computationally intensive phases of the component. We demonstrate its use in improving the performance of one of the most prominent multicomponent applications, the community climate system model (CCSM). In this application, we dynamically extend the atmosphere component to minimize the idling in other components caused by large periodic temporal load imbalances in the atmosphere component. By means of experiments on different parallel platforms with different numbers of processors, we show that using our strategy can lead to about 15% reduction and savings of several days in execution times of CCSM for 1000-year simulation runs.

[1]  James J. Hack,et al.  A VFSA Scheduler for Radiative Transfer Data in Climate Models , 2004, PDCS.

[2]  Nicholas M. Patrikalakis,et al.  Adaptive Coupled Physical and Biogeochemical Ocean Predictions: A Conceptual Basis , 2004, International Conference on Computational Science.

[3]  Ravi S. Nanjundiah.  Strategies for parallel implementation of a global spectral atmospheric general circulation model , 1998, Proceedings. Fifth International Conference on High Performance Computing (Cat. No. 98EX238).

[4]  Arthur A. Mirin,et al.  A Scalable Implementation of a Finite-Volume Dynamical Core in the Community Atmosphere Model , 2005, Int. J. High Perform. Comput. Appl..

[5]  Ravi S. Nanjundiah,et al.  Computational load in model physics of the parallel NCAR community climate model , 1994 .

[6]  Ian Foster,et al.  Load-balancing algorithms for climate models , 1994, Proceedings of IEEE Scalable High Performance Computing Conference.

[7]  Quentin F. Stout,et al.  A PHYSICS-BASED SOFTWARE FRAMEWORK FOR SUN-EARTH CONNECTION MODELING , 2005 .

[8]  Jay Walter Larson,et al.  The Model Coupling Toolkit: A New Fortran90 Toolkit for Building Multiphysics Parallel Coupled Models , 2005, Int. J. High Perform. Comput. Appl..

[9]  John B. Drake,et al.  Porting and Performance of the Community Climate System Model (CCSM3) on the Cray X1 , 2005 .

[10]  Ian T. Foster,et al.  Design and Performance of a Scalable Parallel Community Climate Model , 1995, Parallel Comput..

[11]  Cecelia DeLuca,et al.  Design and Implementation of Components in the Earth System Modeling Framework , 2005, Int. J. High Perform. Comput. Appl..

[12]  W. Collins,et al.  The Community Climate System Model: CCSM3 , 2004 .

[13]  Charles Lawrence.  An Overview of Three Approaches to Multidisciplinary Aeropropulsion Simulations , 1997 .

[14]  Patrick H. Worley,et al.  Extending Scalability of the Community Atmosphere Model , 2007 .

[15]  Gita Alaghband,et al.  The Promise of Load Balancing the Parameterization of Moist Convection Using a Model Data Load Index , 2006 .

[16]  Rafael Delgado-Buscalioni,et al.  Hybrid molecular-continuum fluid models: implementation within a general coupling framework , 2005, Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[17]  K. Taylor,et al.  The Community Climate System Model , 2001 .

[18]  Ravi S. Nanjundiah,et al.  Parallel implementation, validation, and performance of MM5 , 1994 .

[19]  J. Larson.  Ten organising principles for coupling in multiphysics and multiscale models , 2009 .

[20]  Rupert W. Ford,et al.  Load Balancing Physics Routines , 1998 .

[21]  Kerry A. Emanuel,et al.  Numerical Instability Resulting from Infrequent Calculation of Radiative Heating , 2004 .

[22]  Philippe Olivier Alexandre Navaux,et al.  Dynamic load balancing in PC clusters: an application to a multiphysics model , 2003, Proceedings. 15th Symposium on Computer Architecture and High Performance Computing.

[23]  Jaideep Ray,et al.  A component-based scientific toolkit for reacting flows , 2003 .

[24]  Jimy Dudhia,et al.  The Weather Research and Forecast Model: software architecture and performance [presentation] , 2005 .

[25]  P. V. Coveney,et al.  On steering coupled models , 2005 .

[26]  Ravi S. Nanjundiah.  Seasonal simulation of the monsoon with the NCMRWF model , 2000 .