This Article 
 Bibliographic References 
 Add to: 
Parallel Processing of Adaptive Meshes with Load Balancing
December 2001 (vol. 12 no. 12)
pp. 1269-1280

Many scientific applications involve grids that lack a uniform underlying structure. These applications are often also dynamic in nature in that the grid structure significantly changes between successive phases of execution. In parallel computing environments, mesh adaptation of unstructured grids through selective refinement/coarsening has proven to be an effective approach. However, achieving load balance while minimizing interprocessor communication and redistribution costs is a difficult problem. Traditional dynamic load balancers are mostly inadequate because they lack a global view of system loads across processors. In this paper, we propose a novel and general-purpose load balancer that utilizes symmetric broadcast networks (SBN) as the underlying communication topology and compare its performance with a successful global load balancing environment, called PLUM, specifically created to handle adaptive unstructured applications. Our experimental results on an IBM SP2 demonstrate that the SBN-based load balancer achieves lower redistribution costs than that under PLUM by overlapping processing and data migration.

[1] C.J. Alpert and A.B. Kahng, “Recent Directions in Netlist Partitioning: A Survey,” VLSI J., vol. 19, nos. 1-2, pp. 1–81, 1995.
[2] R. Biswas and L. Oliker, “Experiments with Repartitioning and Load Balancing Adaptive Meshes,” Grid Generation and Adaptive Algorithms, vol. 113, pp. 89-111, 1999.
[3] N. Chrisochoides, “Multithreaded Model for the Dynamic Load Balancing of Parallel Adaptive PDE Computations,” Applied Numerical Math., vol. 20, pp. 349-365, 1996.
[4] S.K. Das and D.J. Harvey, “Performance Analysis of an Adaptive Symmetric Broadcast Load Balancing Algorithm on the Hypercube,” Technical Report CRPDC-95-1, Dept. of Computer Science, Univ. of North Texas, 1995.
[5] S.K. Das, D.J. Harvey, and R. Biswas, “Adaptive Load Balancing Algorithms Using Symmetric Broadcast Networks: Performance Study on an IBM SP2,” Proc. 26th Int'l Conf. Parallel Processing, pp. 360-367, 1997.
[6] S.K. Das and S.K. Prasad, “Implementing Task Ready Queues in a Multiprocessing Environment,” Proc. Int'l Conf. Parallel Computing, pp. 132-140, 1990.
[7] S.K. Das, S.K. Prasad, C-Q. Yang, and N.M. Leung, “Symmetric Broadcast Networks for Implementing Global Task Queues and Load Balancing in a Multiprocessor Environment,” Technical Report CRPDC-92-1, Dept. of Computer Science, Univ. of North Texas, 1992.
[8] B. Hendrickson and R. Leland, “An Multilevel Algorithm for Partitioning Graphs,” Proc. Supercomputing, Dec. 1995
[9] G. Horton, “A Multi-Level Diffusion Method for Dynamic Load Balancing,” Parallel Computing, vol. 19, pp. 209-229, 1993.
[10] Y.F. Hu and R.J. Blake, “An Optimal Dynamic Load Balancing Algorithm,” Technical Report DL-P-95-0-11, Daresbury Laboratory, 1995.
[11] G. Karypis and V. Kumar, “Parallel Multilevel k-Way Partitioning Scheme for Irregular Graphs,” Technical Report 96-036, Dept. of Computer Science, Univ. of Minnesota, 1996.
[12] B.W. Kernighan and S. Lin, “An Efficient Heuristic Procedure for Partitioning Graphs,” Bell Systems Technical J., vol. 49, pp. 291-307, 1970.
[13] G. Kohring, "Dynamic Load Balancing for Parallelized Particle Simulations on MIMD Computers," Parallel Computing, vol. 21, pp. 683-693, 1995.
[14] N. Mansour, R. Ponnusamy, A.N. Choudhary, and G.C. Fox, “Graph Contraction and Physical Optimization Methods: A Quality-Cost Tradeoff for Mapping Data on Parallel Computers,” Proc. Seventh Int'l Conf. Supercomputing, pp. 1-10, 1993.
[15] L. Oliker and R. Biswas, “PLUM: Parallel Load Balancing for Adaptive Unstructured Meshes,” J. Parallel and Distributed Computing, vol. 52, pp. 150–177, 1998.
[16] L. Oliker, R. Biswas, and H.N. Gabow, “Parallel Tetrahedral Mesh Adaptation with Dynamic Load Balancing,” Parallel Computing, vol. 26, pp. 1583-1608, 2000.
[17] K. Schloegel, G. Karypis, and V. Kumar, “Multilevel Diffusion Schemes for Repartitioning of Adaptive Meshes,” J. Parallel and Distributed Computing, vol. 47, pp. 109-124, 1997.
[18] H. Shan, J.P. Singh, L. Oliker, and R. Biswas, “A Comparison of Three Programming Models for Adaptive Applications on the Origin2000,” Proc. Supercomputing '00, 2000.
[19] R. Van Driessche and D. Roose, “Load Balancing Computational Fluid Dynamics Calculations on Unstructured Grids,” Parallel Computing in CFD, AGARD-R-807, pp. 2.1-2.26, 1995.
[20] A. Vidwans, Y. Kallinderis, and V. Venkatakrishnan, “Parallel Dynamic Load Balancing Algorithm for Three-Dimensional Adaptive Unstructured Grids,” American Inst. Aeronautics and Astronautics J., vol. 32, pp. 495-505, 1994.
[21] C. Walshaw, M. Cross, and M.G. Everett, “Parallel Dynamic Graph Partitioning for Adaptive Unstructured Meshes,” J. Parallel and Distributed Computing, vol. 47, pp. 102-108, 1997.

Index Terms:
Dynamic load balancing, experimental study, IBM SP2, job migration and redistribution, symmetric broadcast networks, unstructured mesh adaptation
S.K. Das, D.J. Harvey, R. Biswas, "Parallel Processing of Adaptive Meshes with Load Balancing," IEEE Transactions on Parallel and Distributed Systems, vol. 12, no. 12, pp. 1269-1280, Dec. 2001, doi:10.1109/71.970562
Usage of this product signifies your acceptance of the Terms of Use.