Abstract
We investigate the lattice-based array partitioning based on the theory of the Smith Normal Form and we present two elegant techniques for partitioning arrays in parallel DoAll loops for message-passing parallel machines: (1) DoAll loops with constant dependencies for communication-free partitioning: a general solution of all possible communication-free partitioning is derived where the dependencies among array references are described in constant distance vectors. (2) DoAll loops with non-constant dependencies for block-communication partitioning: the dependencies among array references are described in non-constant distance vectors. We derive the partitioning equations which allocate all remote data to a unique processor such that only one block-communication can obtain all the remote data for the computation. By using the Smith Normal Form decomposition, we are also able to verify our partitioning results.
Similar content being viewed by others
References
E. H. D’Hollander (1992) ArticleTitlePartitioning and Labeling of Loops by Unimodular Transformations IEEE Trans. Parallel Distrib. Syst. 3 IssueID4 465–476 Occurrence Handle10.1109/71.149964
H. Wijshoff and J. Leeuwen, The Structure of Periodic Storage Schemes for Parallel Memories, IEEE Trans. Comp. c-34(6) (June 1985).
E. H.- Y. Tseng J.-L. Gaudiot (1999) ArticleTitleCommunication Generation for Aligned and Cyclic (K) Distributions Using Integer Lattice IEEE Trans. Parallel Distrib. Syst. 10 IssueID2 136–146 Occurrence Handle10.1109/71.752780
J. Ramanujam P. Sadayappan (1991) ArticleTitleCompile-Time Techniques for Data Distribution in Distributed Memory Machines IEEE Trans. Parallel Distrib. Syst. 2 IssueID4 472–482 Occurrence Handle10.1109/71.97903
M. Mace (1987) Memory Storage Pattern in Parallel Processing Kluwer Dordrecht
M. Dion Y. Robert (1996) ArticleTitleMapping affine loop nests Parallel Comput. 22 IssueID10 1373–1397 Occurrence Handle10.1016/S0167-8191(96)00049-X
P. Feautrier, Automatic Distribution of Data and Computations, PRiSM (March 2000).
A. Lim and M. Lam, Communication-Free Parallelization via Affine Transformations, in Proc.7th workshop on languages and compilers for parallel computing (1994).
J. Ramanujam, Non-unimodular transformations of nested loops, Proceedings of the 1992 ACM/IEEE conference on Supercomputing, IEEE Computer Society Press,\break pp. 214–223, (1992).
M. Wolfe (1995) High Performance Compilers for Parallel Computing Addison-Wesley Reading, MA
P. Feautrier (1992) ArticleTitleSome efficient solutions to the affine scheduling problem I. One-dimensional time Int. J. Parallel Program 21 IssueID5 313–348
M. Gupta and P. Banerjee, PARADIGM: A compiler for automatic data distribution on multicomputers, Proceedings of the 7th international conference on Supercomputing, ACM Press, New York, pp. 87–96, (1993).
K. Kennedy U. Kremer (1998) ArticleTitleAutomatic data layout for distributed-memory machines ACM Trans. Program Lang. Syst. 20 IssueID4 869–916 Occurrence Handle10.1145/291891.291901
A. Thirumalai J. Ramanujam (1996) ArticleTitleEfficient computation of address sequences in data parallel programs using closed forms for basis vectors J. Parallel Distrib. Comput. 38 IssueID2 188–203 Occurrence Handle10.1006/jpdc.1996.0140
C. Ancourt F. Coelho F. Irigoin R. Keryell (1997) ArticleTitleA linear algebra framework for static High Performance Fortran code distribution Sci. Program 6 IssueID1 3–27
K. Reeuwijk Particlevan W. Denissen H. J. Sips E. M. R. M. Paalvast (1996) ArticleTitleAn Implementation Framework for HPF Distributed Arrays on Message-Passing Parallel Computer Systems IEEE Trans. Parallel Distrib. Syst. 7 IssueID9 897–914 Occurrence Handle10.1109/71.536935
S. Chatterjee J. R. Gilbert F. J. E. Long R. Schreiber S.-H. Teng (1995) ArticleTitleGenerating local addresses and communication sets for data-parallel programs J. Parallel Distrib. Comput. 26 IssueID1 72–84 Occurrence Handle10.1006/jpdc.1995.1049
M. Newman (1972) Integral Matrices Academic Press New York
D. Bau, I. Kodukula, V. Kotlyar, K. Pingali, and P. Stodghill, Solving Alignment Using EIementary Linear Algebra, Proceedings of the 7th International Workshop on Languages and Compilers for Parallel Computing, Springer-Verlag, Berlin, pp. 46–60, (1994).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Tseng, EY., Gaudiot, JL. Automatic Array Partitioning Based on the Smith Normal Form. Int J Parallel Prog 33, 35–56 (2005). https://doi.org/10.1007/s10766-004-1460-2
Received:
Revised:
Accepted:
Issue Date:
DOI: https://doi.org/10.1007/s10766-004-1460-2