A Repeated Mapping Scheme of Task Modules with Minimum Communication Cost in Hypercube Multicomputers

  • Published : 1998.12.15

Abstract

This paper deals with the problem of one-to-one mapping of 2$^n$ task modules of a parallel program to an n-dimensional hypercube multicomputer so as to minimize the total communication cost during the execution of the task. The problem of finding an optimal mapping has been proven to be NP-complete. First we show that the mapping problem in a hypercube multicomputer can be transformed into the problem of finding a set of maximum cutsets on a given task graph using a graph modification technique. Then we propose a repeated mapping scheme, using an existing graph bipartitioning algorithm, for the effective mapping of task modules onto the processors of a hypercube multicomputer. The repeated mapping scheme is shown to be highly effective on a number of test task graphs; it increasingly outperforms the greedy and recursive mapping algorithms as the number of processors increases. Our repeated mapping scheme is shown to be very effective for regular graphs, such as hypercube-isomorphic or 'almost' isomorphic graphs and meshes; it finds optimal mappings on almost all the regular task graphs considered.

Keywords

References

  1. Proc. of the Second Conf. on Hypercube Concurrent Computers and Applications Embedding of Interacting Task Modules into a Hypercube Chen, M.S.;Shin, K.G.
  2. Proc. of 27th Annual Symp. on Foundations Computer Sci How Robust is the n-cube Becker, B.;Simon, H.U.
  3. Proc. of the 1989 Conf. on Parallel Processing A Processor Allocation Scheme for Hypercube Computers Kim, J.;Das, C.R.;Lin, W.
  4. Proc. of AFIPS Conf. v.54 Concurrent Processing: A New Direction in Scientific Computing Rattner, J.
  5. Proc. of the 1986 Conf. on Parallel Processing Architecture of a Hypercube Supercomputer Hayes, J.P.;Mudge, T.N.(et al.)
  6. Proc. of the 1985 Conf. on Parallel Processing The Mark III Hypercube Ensemble Concurrent Processor Peterson, J.C.(et al.)
  7. The Connection Machine Hills, W.D.
  8. IEEE Trans. on Computers v.C-30 no.3 On the Mapping Problem Bokhari, S.H.
  9. IEEE Trans. on Computers v.C-36 no.12 Processor Allocation in an n-cube Multiprocessor Using Gray Codes Chen, M.S.;Shin, K.G.
  10. IEEE Trans. on Computers v.C-33 no.9 Wave Scheduling-Decentralized Scheduling of Task Forces in Multicomputers van Tilborg, A.M.;Wittie, L.D.
  11. Proc. of the First Conf. on Hypercube Concurrent Computers and Applications Hypercube Embedding is NP-Complete Krumme, D.W.;Venkataraman, K.N.;Cybenko, G.
  12. Proc. of the Third Conf. on Hypercube Concurrent Computers and Applications A Graphoriented Mapping Strategy for a Hypercube Chen, W.K.;Gehringer, E.F.
  13. Proc. of the Third Conf. on Hypercube Concurrent Computers and Applications Task Allocation onto a Hypercube by Recursive Mincut Bipartitioning Ercal, F.;Ramanujan, J.;Sadayappan, P.
  14. Bell Syst. Tech. J. v.49 An Efficient Heuristic Procedure for Partitioning Graphs Kernighan, B.W.;Lin, S.
  15. Computer-Aided Design v.21 no.10 Efficient Algorithm for Graph-Partitioning Problem Using a Problem Transformation Method Lee, C.H.;Park, C.I.;Kim, M.
  16. Proc. of 12-th Int'l Conf. on Distributed Comput. Syst. Communication-oriented Assignment of Task Modules in Hypercube Multicomputers Tsai, B.R.;Shin, K.G.
  17. Comput. Networks v.3 Virtual Cut-through: A New Computer Communication Switching Technique Kermani, P.;Kleinrock, L.
  18. IEEE Trans. on Software Engineering v.SE-3 no.1 Multiprocessor Scheduling with the Aid of Network Flow Algorithms Stone, H.S.
  19. IEEE Trans. on Computers v.C-41 no.7 Optimal Task Assignment in Linear Array Networks Lee, C.H.;Lee, D.;Kim, M.
  20. IEEE Trans. on Parallel and Distributed Systems v.8 no.2 Optimal Task Assignment in Homogeneous Systems Lee, C.H.;Shin, Kang G.