Abstract
We present Natural Load Indices (NLIs) as an alternative to measurement-based load indices. NLIs facilitate further performance improvement and better resource usage. Example NLIs are rainfall amounts in a climate simulation, mass of an atom in a Molecular Dynamics (MD) code and surface fluxes in an ocean model. The process of obtaining an NLI occurs during model development or as a preprocessing step and implementing NLIs minimizes run-time costs associated with dynamic load balancing.
NLIs are investigated in several applications. (A) We implement NLIs in the Community Atmosphere Model and discuss performance. (B) We extend prior Molecular Dynamics work by providing performance analysis using longer simulations. (C) We demonstrate the existence of an NLI in an ocean model. (D) We demonstrate similarities between NLIs in those models and show that NLIs are a general class of load index. Systems investigated include Pentium 4 Xeon, IBM Power5-p575 and IBM BlueGene/L parallel platforms.
Similar content being viewed by others
References
Amundson J, Spentzouris P, Qiang J, Ryne R (2006) Synergia: an accelerator modeling tool with 3-D space charge. J Comput Phys 211(1):229–248
Behrens J (1997) Atmospheric and ocean modeling with an adaptive finite element solver for the shallow-water equations. Trans IMACS Appl Numer Math 26(1–2):217–226
Bhandarkar M, Kalé L, Sturler E, Hoeflinger J (2000) Object-based adaptive load balancing for mpi programs
Blayo E, Debreu L, Mounie G, Trystram D (1999) Dynamic load balancing for ocean circulation model with adaptive meshing. In: Euro-Par ’99: proceedings of the 5th international Euro-par conference on parallel processing, London, UK. Springer, Berlin, pp 303–312
BlueGene/l information (2010) Technical Report. https://wiki.cs.colorado.edu/bluegenewiki/frontpage
Bluevista information (2010) Technical Report. http://www.cisl.ucar.edu/computers/bluevista/
Bull RWFJM, Dickinson A (1996) A feedback based load balance algorithm for physics routines in NWP. In: Proceedings of seventh ECMWF workshop on the use of parallel processors in meteorology. World Scientific, Singapore
Collins WD, et al (2004) Description of the NCAR Community Atmosphere Model (CAM 3.0). Technical Report NCAR/TN-464+STR, NCAR, National Center for Atmospheric Research, Boulder, CO
Deng Y, McCoy R, Marr R, Peierls R, Yasar O (1995) Molecular dynamics on distributed-memory MIMD computers with load balancing. Appl Math Lett 8:37–41
Dent D, Isaksen L, Mozdzynski G, Robinson G, Wollenweber F, O’Keefe M (1995) The IFS Model performance measurements. In: Hoffmen GR, Kreitz N (eds) Coming of age, Proceedings of the 6th ECMWF on the use of parallel processors in meteorology. World Scientific, Singapore, pp 252–369
Devine KD, Boman EG, Heaphy R, Hendrickson B, Teresco JFJD, Flaherty J, Gervasio L (2004) New challanges in dynamic load balancing. Appl Numer Math 52:133–152
Drake J, Flanery RE, Smeraro BD, Worley PH, Foster IT, Hack JJ, Williamson DL (1996) Parallel community climate model: description and users guide. Technical Report ORNL/TM-12285, Oak Ridge National Laboratory
Ferrari D, Zhou S (1987) An empirical investigation of load indices for load balancing applications. In: Performance ’87, December 1987
Ford RW, Burton PM (1998) Load balancing physics routines. In: Proceedings of the eighth ECMWF workshop on the use of parallel processors in meteorology, November 1998
Ford RW, Snelling DF, Dickinson A (1994) Controlling load balance, cache use and vector length in the UM. In: Sixth workshop on use of parallel processors in meteorology at ECMWF, November1994
Foster IT, Toonen BR (1994) Load-balancing algorithms for climate models. In: Proc scalable high performance computing conf. IEEE Computer Society, Los Alamitos, pp 674–681
Hack JJ (1994) Parameterization of moist convection in the National Center for Atmospheric Research community climate model (CCM2). J Geophys Res 99(D3):5541–5568
Hayashi R, Horiguchi S (2000) Efficiency of dynamic load balancing based on permanent cells for parallel molecular dynamics simulation. In: IPDPS, pp 85–92
Hayashi R, Horiguchi S (2000) Relationships between efficiency of dynamic load balancing and particle concentration for parallel molecular dynamics simulation. In: Proc of high performance computing Asia, pp 976–983
Jones PW, Worley PH, Yoshida Y, White JB III, Levesque J (2005) Practical performance portability in the parallel ocean program (pop): research articles. Concurr Comput, Pract Exp 17(10):1317–1327
Kalé LV, Bhandarkar M, Brunner R (1998) Load balancing in parallel molecular dynamics. In: Fifth international symposium on solving irregularly structured problems in parallel Lecture notes in computer science, vol 1457. Springer, Berlin, pp 251–261
Kalé L, Skeel R, Bhandarkar M, Brunner R, Gursoy A, Krawetz N, Phillips J, Shinozaki A, Varadarajan K, Schulten K (1999) Namd2: greater scalability for parallel molecular dynamics. J Comput Phys 151(1):283–312
Kantha LH (2005) Development, testing and implementation of a real-time nowcast/forecast capability for the gulf of Mexico. Kaiyo Monthly 37(4):239–256
Kantha LH, Clayson CA (2000) Numerical models of oceans and oceanic processes. International geophysics series, vol 66. Academic Press, San Diego
Kerbyson DJ, Jones PW (2005) A performance model of the parallel ocean program. Int J High Perform Comput Appl 19(3):261–276
Knecht R, Kohring GA (2001) Dynamic load balancing for the simulation of granular materials. In: ICS ’95: proceedings of the 9th international conference on supercomputing. ACM Press, New York, pp 164–169
Kunz T (1991) The influence of different workload descriptions on a heuristic load balancing scheme. IEEE Tran Softw Eng 17(7):725–730
Michalakes JG (1991) Analysis of workload and load balancing issues in the NCAR community climate model. Technical Report ANL/MCS-TM-144, Argonne National Laboratory, Argonne, IL
Michalakes JG, Nanjundiah RS (1994) Computational load in model physics of the parallel NCAR community climate model. Technical Report ANL/MCS-TM-186, Argonne National Laboratory, Argonne, IL
Mpi reference (2010) Technical Report. http://www.mpi-forum.org
Muszala SP, Alaghband G, Connors DA, Hack JJ (2004) A VFSA scheduler for radiative transfer data in climate models. In: 17th international conference on parallel and distributed computing systems, September 2004
Muszala SP, Hack JJ, Connors DA, Alaghband G (2006) The promise of load balancing the parameterization of moist convection using a model data load index. J Atmos Ocean Technol 23:525–537
Muszala S, Alaghband G, Hack J, Connors DA (2008) Natural Load Indices (NLI) in NAMD2 load balancing algorithms, corresponding LNCS paper in review. In: PARA 2008, 9th international workshop on state-of-the-art in scientific and parallel computing, May 2008
Nakano A, Campbell T (1997) An adaptive curvilinear-coordinate approach to dynamic load balancing of parallel multiresolution molecular dynamics. Parallel Comput 23(10):1461–1478
Nanjundiah RS (2000) Seasonal simulation of the monsoon with the NCMRWF model. Curr Sci 78:869–875
Nieter C, Cary JR (2004) Vorpal: a versatile plasma simulation code. J Comput Phys 196(2):448–473
Nyland LS, Prins J, Yun RH, Hermans J, Kum H-C, Wang L (1998) Modeling dynamic load balancing in molecular dynamics to achieve scalable parallel execution. In: IRREGULAR ’98: Proceedings of the 5th international symposium on solving irregularly structured problems in parallel, London, UK. Springer, Berlin, pp 356–365
Phillips J, Zheng G, Kalé LV (2002) Namd: biomolecular simulation on thousands of processors. In: Workshop: scaling to new heights, Pittsburgh, PA, May 2002
Phillips JC, Braun R, Wang W, Gumbart J, Tajkhorshid E, Villa E, Chipot C, Skeel RD, Kalé L, Schulten K (2005) Scalable molecular dynamics with namd. J Comput Chem 26:1781–1802
Pop web site url (2010) Technical Report. http://climate.lanl.gov/models/pop/
Pop user guide url (2010) Technical Report. http://climate.lanl.gov/models/pop/documentation
Purohit S, Kaginalkar A, Jindani I, Ratnam JV, Dash SK (1999) Development of parallel climate/forecast models on 100 GFlops PARAM computing systems. In: Proceedings of the eight ECMWF workshop on the use of parallel processors in meteorology
Rosinski JM, Williamson DL (1997) The accumulation of rounding errors and port validation for global atmospheric models. SIAM J Sci Comput 18(2):552–564
Sandia (2009) LAMMPS benchmarks. Technical Report
Serio AD, Ibáñez MB (2002) Distributed load balancing for molecular dynamics simulations. In: HPCS, pp 283–288
Society AM (2006) AMS glossary of meteorology. http://amsglossary.allenpress.com/
Straatsma TP, McCammon JA (2001) Load balancing of molecular dynamics simulation with nwchem. IBM Syst J 40(2):328–341
Zhang GJ, McFarlane NA (1995) Sensitivity of climate simulations to the parameterization of cumulus convection in the Canadian Climate Centre general circulation model. Atmos Ocean 33:407–446
Xu C, Lau F (1997) Load balancing in parallel computers, theory and practice. Kluwer Academic, Dordrecht
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Muszala, S.P., Alaghband, G., Hack, J. et al. Natural Load Indices (NLI) for scientific simulation. J Supercomput 59, 392–413 (2012). https://doi.org/10.1007/s11227-010-0442-y
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11227-010-0442-y