Cyclo-Static Realizations, Loop Unrolling and CPM: Optimal Multiprocessor Scheduling

  • D. A. Schwartz


Optimal, deterministic scheduling, for a class of iterative/recursive problems is explored. The connection between the program transformation technique of loop unrolling and the blocking of a related acyclic graph for the determination of CPM schedules is developed. It is shown that loop unrolling and blocking can increase the parallelism of the realization by allowing the overlapped execution of successive iterations. Optimal overlapping is based on simple bounds for a related cyclic graph. Cyclo-static realizations that can achieve the optimal overlapped schedules are then introduced.


Directed Acyclic Graph Critical Path Successive Iteration Slack Time Atomic Operation 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    D. A. Schwartz and T. P. Barnwell III, “Cyclo-Static Multiprocessor Scheduling for the Optimal Implementation of Shift-Invariant Flow Graphs,” ICASSP’85, Tampa, FL, March 1985.Google Scholar
  2. [2]
    D. A. Schwartz, “Synchronous Multiprocessor Realizations of Shift Invariant Flow Graphs,” Elec. Eng., Georgia Inst. of Tech, DSPL-85–2, July, 1985.Google Scholar
  3. [3]
    D. A. Schwartz and T. P. Barnwell III, “Cyclo-Static Solutions: Optimal Multiprocessor Realizations of Recursive Algorithms,” VLSI Signal Processing II, Chap. 11, IEEE Press, 1986.Google Scholar
  4. [4]
    M. J. Gonzalez, “Deterministic Processor Scheduling,” Computing Surveys, vol. 9, no. 3, Sept. 1977, pp. 173–204.CrossRefGoogle Scholar
  5. [5]
    E. G. Coffman, ed. Computer and Job-Shop Scheduling Theory, John Wiley, New York, 1976.Google Scholar
  6. [6]
    J. Ondás, “Algorithms for Scheduling Homogeneous Multiprocessor Computers,” Algorithms, Software and Hardware of Parallel Computers, Eds. J. Miklosko and V. E. Kotov, Springer-Verlag, 1984.Google Scholar
  7. [7]
    Jan Zeman and G. S. Moschytz, “Systematic Design and Programming of Signal Processors, Using Project Management Techniques,” IEEE Trans. on ASSP, Dec. 1983, pp. 1536–1549.Google Scholar
  8. [8]
    E. A. Lee, “A Coupled Hardware and Software Architecture for Programmable Digital Signal Processors,” Memorandum No. UCB/ERL M86/54, ERL, Univ. of California, Berkeley, June, 1986.Google Scholar
  9. [9]
    Markku Renfors and Yrjö Neuvo, “The Maximum Sampling Rate of Digital Filters Under Hardware Speed Constraints,” IEEE Trans. on Cir. and Sys., March 1981, pp. 196–202.Google Scholar
  10. [10]
    A. Fettweis, “Realizability of Digital Filter Networks,” Arch. Elek. Übertragung., Feb. 1976, pp. 90–96.Google Scholar
  11. [11]
    T. P. Barnwell III, C. J. M. Hodges, M. Randolf, “Optimal Implementation of Single Time Index Signal Flow Graphs on Synchronous Multiprocessor Machines,” ICASSP’82, Paris, France, May, 1982.Google Scholar
  12. [12]
    C. E. Leiserson, “Optimizing Synchronous Circuitry by Retiming,” in Proc. 3rd Caltech Conf. Very Large Scale Integration, ed. R. Bryant, Computer Science Press, 1983, pp. 87–116.Google Scholar
  13. [13]
    H. G. Forren, personal communication, 1987.Google Scholar
  14. [14]
    H. G. Forren and D. A. Schwartz, “Transforming Periodic Synchronous Multiprocessor Programs,” Proc. ICASSP’87, Houston, TX, April, 1987.Google Scholar
  15. [15]
    D. A. Schwartz, T. P. Barnwell III, and C. J. M. Hodges, “The Optimal Synchronous Cyclo-static Array: A Multiprocessor Supercomputer for Digital Signal Processing,” Proc. ICASSP’86, Tokyo, Japan, April, 1986.Google Scholar
  16. [16]
    S. J. A. McGrath, T. P. Barnwell III, and D. A. Schwartz, “A WE-DSP32 Based, Low-Cost, High Performance, Synchronous Multiprocessor for CycloStatic Implementations,” Proc. ICASSP’87, Houston, TX, April, 1987.Google Scholar

Copyright information

© Plenum Press, New York 1988

Authors and Affiliations

  • D. A. Schwartz
    • 1
  1. 1.School of Electrical EngineeringGeorgia Institute of TechnologyUSA

Personalised recommendations