Analyzing the impact of computational heterogeneity on runtime performance of parallel scientific components

ABSTRACT

Scientific simulations modeling complex physical phenomena exhibit varying degrees of spatiotemporal and computational heterogeneity, which can pose significant challenges in algorithmic efficiency and runtime performance. Addressing these challenges often requires an understanding of application behavior and the impact of heterogeneity on the simulation, especially when analytical approaches are not feasible. Runtime calibration is used to analyze the impact of computational heterogeneity on the performance of two different load balancing strategies applied to two different problems of 2-D methane-air combustion using different chemistry models. Experimental evaluation demonstrates that such empirical approaches are inevitable in component-based scientific computations, where performance is determined by the problem characteristics and the particular connectivity of components in the simulation code, none of which are known before runtime.

REFERENCES

• 1: Sumir Chandra , Shweta Sinha , Manish Parashar , Yeliang Zhang , Jingmei Yang , Salim Hariri, Adaptive Runtime Managementof SAMR Applications, Proceedings of the 9th International Conference on High Performance Computing, p.564-574, December 18-21, 2002
• 2: M. J. Berger , P. Colella, Local adaptive mesh refinement for shock hydrodynamics, Journal of Computational Physics, v.82 n.1, p.64-84, May 1989 [doi>10.1016/0021-9991(89)90035-1]
• 3: S. Bhowmick , P. Raghavan , L. McInnes , B. Norris, Faster PDE-based simulations using robust composite linear solvers, Future Generation Computer Systems, v.20 n.3, p.373-387, April 2004 [doi>10.1016/j.future.2003.07.012]
• 4: Bhowmick, S., L. C. McInnes, B. Norris, and P. Raghavan. 2003. "The Role of Multi-Method Linear Solvers in PDE-based Simulations." Computational Science and its Applications - ICCSA 2003, Lecture Notes in Computer Science, vol. 2667, 828--839.
• 5: McInnes, L. C., B. Norris, S. Bhowmick, and P. Raghavan. 2003. "Adaptive Sparse Linear Solvers for Implicit CFD using Newton-Krylov Algorithms." In Proceedings of the 2nd MIT Conference on Computational Fluid and Solid Mechanics, Cambridge, MA, Elsevier, vol. 2, 1024--1028.
• 6: Lefantzi, S., J. Ray, C. A. Kennedy, and H. N. Najm. 2005. "A Component-based Toolkit for Reacting Flows with High Order Spatial Discretizations on Structured Adaptively Refined Meshes." Progress in Computational Fluid Dynamics, vol. 5, no. 6, 298--315.
• 7: Kenny, J. P., S. B. Benson, et al. 2004. "Component-based Integration of Chemistry and Optimization Software." Journal of Computational Chemistry, vol. 25, 1717--1725.
• 8: Colella, P.; et al. 2005. Chombo Infrastructure for Adaptive Mesh Refinement. http://seesar.lbl.gov/ANAG/-chombo/.
• 9: Manish Parashar , James C. Browne , Carter Edwards , Kenneth Klimkowski, A common data management infrastructure for adaptive algorithms for PDE solutions, Proceedings of the 1997 ACM/IEEE conference on Supercomputing (CDROM), p.1-22, November 15-21, 1997, San Jose, CA [doi>10.1145/509593.509649]
• 10: Kohn, S. 2005. SAMRAI: Structured Adaptive Mesh Refinement Applications Infrastructure. http://-www.llnl.gov/CASC/SAMRAI/.
• 11: Benjamin A. Allan , Robert Armstrong , David E. Bernholdt , Felipe Bertrand , Kenneth Chiu , Tamara L. Dahlgren , Kostadin Damevski , Wael R. Elwasif , Thomas G. W. Epperly , Madhusudhan Govindaraju , Daniel S. Katz , James A. Kohl , Manoj Krishnan , Gary Kumfert , J. Walter Larson , Sophia Lefantzi , Michael J. Lewis , Allen D. Malony , Lois C. Mclnnes , Jarek Nieplocha , Boyana Norris , Steven G. Parker , Jaideep Ray , Sameer Shende , Theresa L. Windus , Shujia Zhou, A Component Architecture for High-Performance Scientific Computing, International Journal of High Performance Computing Applications, v.20 n.2, p.163-202, May 2006 [doi>10.1177/1094342006064488]
• 12: CCA Forum. 2004. Common Component Architecture Forum. http://www.cca-forum.org.
• 13: Najm, H. N., R. W. Schefer, R. B. Milne, C. J. Mueller, K. D. Devine, and S. N. Kempka. 1998. "Numerical and Experimental Investigation of Vortical Flow-Flame Interaction." Sandia Technical Report SAND98-8232, UC-1409, Sandia National Laboratories, Livermore, CA.
• 14: Omar M. Knio , Habib N. Najm , Peter S. Wyckoff, A semi-implicit numerical scheme for reacting flow: II. stiff, operator-split formulation, Journal of Computational Physics, v.154 n.2, p.428-467, Sept. 20, 1999 [doi>10.1006/jcph.1999.6322]
• 15: Sophia Lefantzi , Jaideep Ray , Habib N. Najm, Using the Common Component Architecture to Design High Performance Scientific Simulation Codes, Proceedings of the 17th International Symposium on Parallel and Distributed Processing, p.52.1, April 22-26, 2003
• 16: Scott D. Cohen , Alan C. Hindmarsh, CVODE, a stiff/nonstiff ODE solver in C, Computers in Physics, v.10 n.2, p.138-143, March/April 1996
• 17: Chandra, S., M. Parashar, and J. Ray. 2006. "Dynamic Structured Partitioning for Parallel Scientific Applications with Pointwise Varying Workloads." In Proceedings of the 20th International Parallel and Distributed Processing Symposium, Rhodes Island, Greece, IEEE Computer Society Press.
• 18: Sagan, H. 1994. Space filling curves, Springer-Verlag.
• 19: Frenklach, M., H. Wang, et al. 1995. "GRI-Mech---An Optimized Detailed Chemical Reaction Mechanism for Methane Combustion." Technical Report GRI-95/0058, Gas Research Institute, Chicago, IL.
• 20: NERSC. 2006. Jacquard Home Page. http://www.nersc.gov/nusers/resources/jacquard.
• 21: Paul, P. H. 1997. "DRFM: A New Package for the Evaluation of Gas-Phase-Transport Properties." Sandia Technical Report SAND98-8203, Sandia National Laboratories, Albuquerque, NM.

INDEX TERMS

Keywords:
component-based combustion applications, high performance computing, load balancing, structured grids