Parallelization of Minimum Spanning Tree Algorithms Using Distributed Memory Architectures

Conference paper

Abstract

Finding a minimum spanning tree of a graph is a well known problem in graph theory with many practical applications. We study serial variants of Prim’s and Kruskal’s algorithm and present their parallelization targeting message passing parallel machine with distributed memory. We consider large graphs that can not fit into memory of one process. Experimental results show that Prim’s algorithm is a good choice for dense graphs while Kruskal’s algorithm is better for sparse ones. Poor scalability of Prim’s algorithm comes from its high communication cost while Kruskal’s algorithm showed much better scaling to larger number of processes.

Keywords

Distributed memory Kruskal MPI MST Paralellization Prim 

Notes

Acknowledgements

Authors are partially supported by Ministry of Education, Science, and Technological Development of the Republic of Serbia, through projects no. ON174023: “Intelligent techniques and their integration into wide-spectrum decision support”, and ON171017: “Modeling and numerical simulations of complex many-body systems”, as well as European Commission through FP7 projects PRACE-2IP and PRACE-3IP.

References

  1. 1.
    H. Ahrabian, A. Nowzari-Dalini, Parallel algorithms for minimum spanning tree problem. Int. J. Comput. Math. 79(4), 441–448 (2002)CrossRefMATHMathSciNetGoogle Scholar
  2. 2.
    D.A. Bader, G. Cong, Fast shared-memory algorithms for computing the minimum spanning forest of sparse graphs. J. Parallel Distrib. Comput. 66(11), 1366–1378 (2006)CrossRefMATHGoogle Scholar
  3. 3.
    O. Boruvka, O Jistém Problému Minimálnm (about a certain minimal problem) (in Czech, German summary). Práce Mor. Prrodoved. Spol. v Brne III, vol. 3 (1926)Google Scholar
  4. 4.
    S. Chung, A. Condon, Parallel implementation of borvka’s minimum spanning tree algorithm. in Proceedings of the 10th International Parallel Processing Symposium, IPPS ‘96 (IEEE Computer Society, Washington, DC, 1996), pp. 302–308Google Scholar
  5. 5.
    T.H. Cormen, C. Stein, R.L. Rivest, C.E. Leiserson, Introduction to Algorithms, 2nd edn. (McGraw-Hill Higher Education, Boston, 2001)MATHGoogle Scholar
  6. 6.
    M. Curtis-Maury, X. Ding, C.D. Antonopoulos, D.S. Nikolopoulos, An Evaluation of Openmp on Current and Emerging Multithreaded/Multicore Processors, ed. by M.S. Mueller, B.M. Chapman, B.R. Supinski, A.D. Malony, M. Voss. OpenMP Shared Memory Parallel Programming, vol 4315 (Lecture Notes in Computer Science, Springer Berlin Heidelberg, 2008), pp. 133–144Google Scholar
  7. 7.
    F. Dehne, S. Gtz, Practical Parallel Algorithms for Minimum Spanning Trees, in Workshop on Advances in Parallel and Distributed Systems (1998), pp. 366–371Google Scholar
  8. 8.
    N. Deo, Y.B. Yoo, Parallel algorithms for the minimum spanning tree problem, in Proceedings of the International Conference on Parallel Processing (1981), pp. 188–189Google Scholar
  9. 9.
    Z. Galil, G.F. Italiano, Data structures and algorithms for disjoint set union problems. ACM Comput. Surv. 23(3), 319–344 (1991)CrossRefGoogle Scholar
  10. 10.
    R.G. Gallager, P.A. Humblet, P.M. Spira, A distributed algorithm for minimum-weight spanning trees. ACM Trans. Program. Lang. Syst. 5(1), 66–77 (1983)CrossRefMATHGoogle Scholar
  11. 11.
    E. Gonina, L.V. Kale, Parallel prim’s algorithm on dense graphs with a novel extension, in PPL Technical Report, Oct 2007Google Scholar
  12. 12.
    R.L. Graham, P. Hell, On the history of the minimum spanning tree problem. IEEE Ann. Hist. Comput. 7(1), 43–57 (1985)CrossRefMATHMathSciNetGoogle Scholar
  13. 13.
    A. Grama, G. Karypis, V. Kumar, A. Gupta, Introduction to Parallel Computing, 2nd edn. (Addison Wesley, Reading, 2003)Google Scholar
  14. 14.
    W. Guang-rong, G. Nai-jie, An efficient parallel minimum spanning tree algorithm on message passing parallel machine. J. Softw. 11(7), 889–898 (2000)Google Scholar
  15. 15.
    M. Harris, Optimizing parallel reduction in CUDA. CUDA tips and tricksGoogle Scholar
  16. 16.
    M. Jin, J.W. Baker, Two graph algorithms on an associative computing model, in Proceedings of the International Conference on Parallel and Distributed Processing Techniques and Applications, PDPTA 2007, vol 1, Las Vegas, Nevada, 25–28 June 2007, pp. 271–277Google Scholar
  17. 17.
    I. Katriel, P. Sanders, J.L. Trff, J.L. Tra, A practical minimum spanning tree algorithm using the cycle property, in 11th European Symposium on Algorithms (ESA), vol. 2832 in LNCS (Springer, New York, 2003), pp. 679–690Google Scholar
  18. 18.
    J.B. Kruskal, On the shortest spanning subtree of a graph and the traveling salesman problem. Proc. Am. Math. Soc. 7(1), 48–50 (1956)CrossRefMATHMathSciNetGoogle Scholar
  19. 19.
    V. Lončar, S. Škrbić, A. Balaž, Distributed memory parallel algorithms for minimum spanning trees, in Lecture Notes in Engineering and Computer Science: Proceedings of the World Congress on Engineering 2013, WCE 2013, London, 3–5 July 2013, pp. 1271–1275Google Scholar
  20. 20.
    Nvidia, Nvidia tesla GPU accelerators. Nvidia Tesla Product Datasheet (2012)Google Scholar
  21. 21.
    V. Osipov, P. Sanders, J. Singler, The filter-kruskal minimum spanning tree algorithm, in ALENEX’09 (2009), pp. 52–61Google Scholar
  22. 22.
    D.-Z. Pan, Z.-B. Liu, X.-F. Ding, Q. Zheng, The application of union-find sets in kruskal algorithm, in Proceedings of the 2009 International Conference on Artificial Intelligence and Computational Intelligence (AICI ‘09), vol 2 (IEEE Computer Society, Washington, DC, 2009), pp. 159–162Google Scholar
  23. 23.
    R.C. Prim, Shortest connection networks and some generalizations. Bell Syst. Technol. J. 36, 1389–1401 (1957)CrossRefGoogle Scholar
  24. 24.
    R. Setia, A. Nedunchezhian, S. Balachandran, A new parallel algorithm for minimum spanning tree problem, in Proceedings of the International Conference on High Performance Computing (HiPC) (2009), pp. 1–5Google Scholar
  25. 25.
    S. Škrbić, Scientific Computing Seminar (2013)Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2014

Authors and Affiliations

  • Vladimir Lončar
    • 1
  • Srdjan Škrbić
    • 1
  • Antun Balaž
    • 2
  1. 1.Faculty of ScienceUniversity of Novi SadNovi SadSerbia
  2. 2.Scientific Computing Laboratory, Institute of Physics BelgradeUniversity of BelgradeBelgradeSerbia

Personalised recommendations