This Article 
 Bibliographic References 
 Add to: 
Parallel Cluster Identification for Multidimensional Lattices
November 1997 (vol. 8 no. 11)
pp. 1089-1097

Abstract—The cluster identification problem is a variant of connected component labeling that arises in cluster algorithms for spin models in statistical physics. We present a multidimensional version of Belkhale and Banerjee's Quad algorithm for connected component labeling on distributed memory parallel computers. Our extension abstracts away extraneous spatial connectivity information in more than two dimensions, simplifying implementation for higher dimensionality. We identify two types of locality present in cluster configurations, and present optimizations to exploit locality for better performance. Performance results from 2D, 3D, and 4D Ising model simulations with Swendson-Wang dynamics show that the optimizations improve performance by 20-80 percent.

[1] K.P. Belkhale and P. Banerjee, "Parallel Algorithms for Geometric Connected Component Labeling on a Hypercube Multiprocessor," IEEE Trans. Computers, vol. 41, no. 6, pp. 699-709, June 1992.
[2] S.J. Fink, S.B. Baden, and K. Jansen, "Cluster Identification on a Distributed Memory Multiprocessor," Proc. 1994 Scalable High Performance Computing Conf., pp. 239-246,Knoxville, Tenn., May 1994.
[3] R.H. Swendson and J.-S Wang, "Nonuniversal Critical Dynamics in Monte Carlo Simulations," Physical Review Letters, vol. 58, pp. 86-88, Jan. 1987.
[4] M. Flanigan and P. Tamayo, "A Parallel Cluster Labeling Method for Monte Carlo Dynamics," Int'l J. Modern Physics C, vol. 3, pp. 1,235-1,249, Dec. 1992.
[5] C.F. Baillie and P.D. Coddington, "Cluster Identification Algorithms for Spin Models—Sequential and Parallel," Concurrency: Practice and Experience, vol. 3, pp. 129-144, Apr. 1991.
[6] D.W. Heermann and A.N. Burkitt, "Parallelization of the Ising Model and Its Performance Evaluation," Parallel Computing, vol. 13, pp. 345-357, 1990.
[7] J. Kertesz and D. Stauffer, "Swenson-Wang Dynamics on Large 2D Critical Ising Models," Int'l J. Modern Physics C, vol. 3, pp. 1,275-1,279, Dec. 1992.
[8] J. Apostolakis, P. Coddington, and E. Marinari, "A Multi-Grid Cluster Labeling Scheme," Europhysics Letters, vol. 17, pp. 189-194, Jan. 1992.
[9] R.C. Brower, P. Tamayo, and B. York, "A Parallel Multigrid Algorithm for Percolation Clusters," J. Statistical Physics, vol. 63, no. 1, pp. 73-88, 1991.
[10] G. Barkema and T. MacFarland, "Parallel Simulation of the Ising Model," Physical Review E, vol. 50, pp. 1,623-1,628, Aug. 1994.
[11] R. Hackl, H.G. Matuttis, J.M. Singer, T. Husslein, and I. Morgenstern, "Parallelization of the 2D Swendson-Wang Algorithm," Proc. Workshop Large Scale Computational Physics on Massively Parallel Computers, H.J. Herrmann and F. Karsch, eds., pp. 59-72. World Scientific, June 1993.
[12] A. Choudhary and R. Thakur, "Evaluation of Connected Component Labeling Algorithms on Shared and Distributed Memory Multiprocessors," Proc. Sixth Int'l Parallel Processing Symp., pp. 362-365, Mar. 1992.
[13] B. Falsafi and R. Miller, "Component Labeling Algorithms on an Intel IPSC/2 Hypercube," Proc. Fifth Distributed Memory Computing Conf., vol. 1, pp. 159-164, Apr. 1990.
[14] H. Embrechts, D. Roose, and P. Wambacq, "Component Labeling on a Distributed Memory Multiprocessor," Proc. First European Workshop Hypercube and Distributed Computers, F. Andre and J. Verjus, eds., pp. 5-17,Amsterdam, 1989.
[15] H. Mino, "A Vectorized Algorithm for Cluster Formation in the Swendson-Wang Dynamics," Computer Physics Comm., vol. 66, pp. 25-30, 1991.
[16] M. Bauernfeind, R. Hackl, H.-G. Matuttis, J. Singer, T. Husslein, and I. Morgenstern, "3D Ising Model with Swendson-Wang Dynamics: A Parallel Approach," Physica A, vol. 212, pp. 277-298, Dec. 1994.
[17] G. Parisi, Statistical Field Theory.Reading, Mass.: Addison-Wesley, 1988.
[18] L. Onsager, "Crystal Statistics: I: A Two-Dimensional Model with an Order-Disorder Transition," Physical Review, vol. 65, pp. 117-149, 1944.
[19] S. Klessinger and G. Munster, "Numerical Investigation of the Interface Tension in the Three-Dimensional Ising Model," Nuclear Physics B, vol. 386, no. 3, pp. 701-713, 1992.
[20] K. Jansen, T. Trappenberg, I. Montvay, G. Munster, and U. Wolff, "Broken Phase of the 4-Dimensional Ising Mdoel in a Finite Volume," Nuclear Physics B, vol. 322, pp. 698-720, 1989.
[21] D.W. Heermann,Computer Simulation Methods in Theoretical Physics, Springer-Verlag, New York, 1990.
[22] T.H. Cormen,C.E. Leiserson, and R.L. Rivest,Introduction to Algorithms.Cambridge, Mass.: MIT Press/McGraw-Hill, 1990.
[23] R.C. Gonzalez and P. Wintz, Digital Image Processing.Reading, Mass.: Addison-Wesley, 1977.
[24] S.R. Kohn and S.B. Baden, "A Robust Parallel Programming Model for Dynamic, Non-Uniform Scientific Computation," Proc. 1994 Scalable High Performance Computing Conf., pp. 509-517,Knoxville, Tenn., May 1994.
[25] S.B. Baden, S.R. Kohn, S.M. Figueira, and S.J. Fink, "The LPARX User's Guide v2.0," technical report, Univ. of California at San Diego, Nov. 1994.
[26] Message Passing Interface Forum, MPI: A Message-Passing Interface Standard (v1.0), May 1994.
[27] J.L. Hennessy and D.A. Patterson, Computer Architecture: A Quantitative Approach, Morgan Kaufmann, San Mateo, Calif., 1990.

Index Terms:
Cluster identification, Ising model, connected component labeling, parallel algorithm, Swendson-Wang dynamics.
Stephen J. Fink, Craig Huston, Scott B. Baden, Karl Jansen, "Parallel Cluster Identification for Multidimensional Lattices," IEEE Transactions on Parallel and Distributed Systems, vol. 8, no. 11, pp. 1089-1097, Nov. 1997, doi:10.1109/71.642944
Usage of this product signifies your acceptance of the Terms of Use.