On the Power of reading and writing simultaneously in parallel computations

  • Rolf Niedermeier
  • Peter Rossmanith
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 762)


In the standard model of Cook, Dwork, and Reischuk [1] for the derivation of lower bounds, one computation step of a CREW-PRAM consists of a read phase, internal computation, and a write phase. We investigate the case when one step comprises only internal computation and one communication phase instead of two, but we have either reading or writing and internal computation, and thus allow the mixed occurrence of reading and writing (of different memory cells) at the same time. We show that simultaneous reading and writing saves communication phases. In detail, we show that the computation of the OR-function requires exactly log n “single communication” steps instead of 0.72 log n “double communication” steps on the standard model. We provide a general lower bound of log(deg(f)) in terms of the degree of a function f. We obtain a lower bound of 0.76 log n for the computation of critical functions and present a critical function that can be computed in 0.90 log n steps. Finally we demonstrate a tight correspondence between CROW-PRAM's of the modified form and the decision tree complexity.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    S. A. Cook, C. Dwork, and R. Reischuk. Upper and lower time bounds for parallel random access machines without simultaneous writes. SIAM Journal on Computing, 15(1):87–97, 1986.CrossRefGoogle Scholar
  2. 2.
    M. Dietzfelbinger, M. Kutyłowski, and R. Reischuk. Exact time bounds for computing boolean functions on PRAMs without simultaneous writes. In Proc. of 2d SPAA, pages 125–135, 1990.Google Scholar
  3. 3.
    P. Dymond and W. L. Ruzzo. Parallel RAMs with owned global memory and deterministic language recognition. In Proc. of 13th ICALP, number 226 in LNCS, pages 95–104. Springer-Verlag, 1986.Google Scholar
  4. 4.
    F. E. Fich. The complexity of computation on the parallel random access machine. In J. H. Reif, editor, Synthesis of Parallel Algorithms, chapter 20, pages 843–900. Morgan Kaufmann Publishers, 1993.Google Scholar
  5. 5.
    F. E. Fich and A. Wigderson. Toward understanding exclusive read. SIAM Journal on Computing, 19(4):718–727, 1990.CrossRefGoogle Scholar
  6. 6.
    J. JáJá. An Introduction to Parallel Algorithms. Addison-Wesley, 1992.Google Scholar
  7. 7.
    R. M. Karp and V. Ramachandran. A survey of parallel algorithms for shared-memory machines. In J. van Leeuwen, editor, Algorithms and Complexity, volume A of Handbook of Theoretical Computer Science, chapter 17, pages 869–932. Elsevier, 1990.Google Scholar
  8. 8.
    M. Kutyłowski. Time complexity of boolean functions on CREW PRAMs. SIAM Journal on Computing, 20(5):824–833, 1991.CrossRefGoogle Scholar
  9. 9.
    N. Nisan. CREW PRAMs and decision trees. SIAM Journal on Computing, 20(6):999–1007, 1991.CrossRefGoogle Scholar
  10. 10.
    I. Parberry and P. Yuan Yan. Improved upper and lower time bounds for parallel random access machines without simultaneous writes. SIAM Journal on Computing, 20(1):88–99, 1991.CrossRefGoogle Scholar
  11. 11.
    R. L. Rivest and J. Vuillemin. On recognizing graph properties from adjacency matrices. TCS, 3:371–384, 1976.CrossRefGoogle Scholar
  12. 12.
    R. Smolensky. Algebraic methods in the theory of lower bounds for boolean circuit complexity. In Proc. of 19th STOC, pages 77–82, 1987.Google Scholar
  13. 13.
    I. Wegener. The Complexity of Boolean Functions. Wiley-Teubner, 1987.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1993

Authors and Affiliations

  • Rolf Niedermeier
    • 1
  • Peter Rossmanith
    • 1
  1. 1.Fakultät für InformatikTechnische Universität MünchenMünchenFed. Rep. of Germany

Personalised recommendations