The Journal of Supercomputing

, 50:289 | Cite as

Parallel loop generation and scheduling



Loop tiling is an efficient loop transformation, mainly applied to detect coarse-grained parallelism in loops. It is a difficult task to apply n-dimensional non-rectangular tiles to generate parallel loops. This paper offers an efficient scheme to apply non-rectangular n-dimensional tiles in non-rectangular iteration spaces, to generate parallel loops. In order to exploit wavefront parallelism efficiently, all the tiles with equal sum of coordinates are assumed to reside on the same wavefront. Also, in order to assign parallelepiped tiles on each wavefront to different processors, an improved block scheduling strategy is offered in this paper.


Loop parallelization Wave-front Code generation Loop scheduling 


  1. 1.
    Allen R, Kennedy K (2001) Optimizing compilers for modern architectures. Morgan Kaufmann, Los Altos Google Scholar
  2. 2.
    Athanasaki M, Sotiropoulos A, Tsoukalas G, Koziris N, Tsanakas P (2005) Hyperplane grouping and pipelined schedules: how to execute tiled loops fast on clusters of SMPs. J Supercomput 33:197–226 CrossRefGoogle Scholar
  3. 3.
    Banerjee U (1993) Loop transformations for restructuring compilers the foundations. Kluwer Academic, Dordrecht MATHGoogle Scholar
  4. 4.
    Bik AJC, Wijshoff HAG (1994) Implementation of Fourier–Motzkin elimination. Leiden University, Leiden Google Scholar
  5. 5.
    Darte A, Robert Y, Vivien F (2000) Scheduling and automatic parallelization. Birkhäuser, Basel MATHGoogle Scholar
  6. 6.
    Eisenbeis C, Sogno JC (1992) A general algorithm for data dependency analysis. In: International Conference on Supercomputing, Washington, July 19–23, 1992 Google Scholar
  7. 7.
    Goumas G, Athanasaki M, Koziris N (2002) Code generation methods for tiling transformations. J Inf Sci Eng 18:667–691 MathSciNetGoogle Scholar
  8. 8.
    Goumas G, Sotiropoulos A, Koziris N (2001) Minimizing completion time for loop tiling with computation and communication overlapping. IEEE Google Scholar
  9. 9.
    Kandemir M, Bordawekar R, Choudhary A, Ramanujam J (1997) A unified tiling approach for out-of-core computations Google Scholar
  10. 10.
    Manjikian N, Abdelrahman TS (1996) Scheduling of wavefront parallelism on scalable shared-memory multiprocessors. Department of Electrical and Computer Engineering, University of Toronto, Toronto Google Scholar
  11. 11.
    Miyandashti FJ (1997) Loop uniformization in shared-memory MIMD machine. Master Thesis, Iran University of Science and Technology (in Persian) Google Scholar
  12. 12.
    Parsa S, Lotfi S (2006) A new genetic algorithms for loop tiling. J Supercomput 37:249–269 CrossRefGoogle Scholar
  13. 13.
    Parsa S, Lotfi S (2006) An outline of a loop parallelization approach in multi-dimensional Cartesian space. In: The 6th Conference perspectives of system informatics, 27–30 June 2006 Google Scholar
  14. 14.
    Rastello F, Robert Y (2002) Automatic partitioning of parallel loops with parallelepiped-shaped tiles. IEEE Trans Parallel Distrib Syst 13(5):460–470 CrossRefGoogle Scholar
  15. 15.
    Wolf ME (1989) Iteration space tiling for memory hierarchies. In: Gary Rodrigue, the 3rd conference on parallel processing for scientific computing, December 1989 Google Scholar
  16. 16.
    Wolf ME (1989) More iteration space tiling. In: Supercomputing’88, November 1989 Google Scholar
  17. 17.
    Wolf ME, Lam MS (1991) A loop transformation theory and an algorithm to maximize parallelism. IEEE Trans Parallel Distrib Syst 2(4):452–471 CrossRefGoogle Scholar
  18. 18.
    Zhao Y, Kennedy K (2005) Scalarization using loop alignment and loop skewing. J Supercomput 31:5–46 MATHCrossRefGoogle Scholar
  19. 19.
    Zima H, Chapman B (1991) Super compilers for parallel and vector computers. Addison-Wesley, Reading Google Scholar

Copyright information

© Springer Science+Business Media, LLC 2009

Authors and Affiliations

  1. 1.Computer Science DepartmentUniversity of TabrizTabrizIran
  2. 2.Faculty of Computer EngineeringIran University of Science and TechnologyTehranIran

Personalised recommendations