Exploiting concurrency among tasks in partitionable parallel processing systems

One benefit of partitionable parallel processing systems is their ability to execute multiple, independent tasks simultaneously. Previous work has identified conditions such that, when there are k tasks to be processed, partitioning the system such that all k tasks are processed simultaneously results in a minimum overall execution time. An alternate condition is developed that provides additional insight into the effects of parallelism on execution time. This result, and previous results, however, assume that execution times are data independent. It is shown that data-dependent tasks do not necessarily execute faster when processed simultaneously even if the condition is met. A model is developed that provides for the possible variability of a task's execution time and is used in a new framework to study the problem of finding an optimal mapping for identical, independent data-dependent execution time tasks onto partitionable systems. Extension of this framework to situations where the k tasks are non-identical is discussed.<<ETX>>

[1]  Vishwani D. Agrawal,et al.  Performance estimation in a massively parallel system , 1990, Proceedings SUPERCOMPUTING '90.

[2]  Shahid H. Bokhari,et al.  A Shortest Tree Algorithm for Optimal Assignments Across Space and Time in a Distributed Processor System , 1981, IEEE Transactions on Software Engineering.

[3]  Tse-Yun Feng,et al.  On a Class of Multistage Interconnection Networks , 1980, IEEE Transactions on Computers.

[4]  Daniel P. Siewiorek,et al.  The Influence of Parallel Decomposition Strategies on the Performance of Multiprocessor Systems , 1985, ISCA.

[5]  Prithviraj Banerjee,et al.  An Approximate Algorithm for the Partitionable Independent Task Scheduling Problem , 1990, ICPP.

[6]  Ramesh Krishnamurti,et al.  The Processor Partitioning Problem In Special-Purpose Partitionable Systems , 1994, ICPP.

[7]  Howard Jay Siegel,et al.  Mapping computer-vision-related tasks onto reconfigurable parallel-processing systems , 1992, Computer.

[8]  Kenneth E. Batcher,et al.  Bit-Serial Parallel Processing Systems , 1982, IEEE Transactions on Computers.

[9]  Howard Jay Siegel,et al.  PASM: A Partitionable SIMD/MIMD System for Image Processing and Pattern Recognition , 1981, IEEE Transactions on Computers.

[10]  Trevor Mudge,et al.  Hypercube supercomputers , 1989, Proc. IEEE.

[11]  Dan C. Marinescu,et al.  Communication and Control in SPMD Parallel Numerical Computations , 1990 .

[12]  L. W. Tucker,et al.  Architecture and applications of the Connection Machine , 1988, Computer.

[13]  David M. Nicol,et al.  Optimal Partitioning of Random Programs Across two Processors , 1989, IEEE Trans. Software Eng..

[14]  Howard Jay Siegel,et al.  Interconnection networks for large-scale parallel processing: theory and case studies (2nd ed.) , 1985 .

[15]  R. F. Freund,et al.  Optimal selection theory for superconcurrency , 1989, Proceedings of the 1989 ACM/IEEE Conference on Supercomputing (Supercomputing '89).

[16]  Cauligi S. Raghavendra,et al.  The Gamma Network , 1984, IEEE Transactions on Computers.

[17]  H. T. Kung,et al.  Synchronized and asynchronous parallel algorithms for multiprocessors , 1976 .

[18]  Alan Weiss,et al.  Allocating Independent Subtasks on Parallel Processors , 1985, IEEE Transactions on Software Engineering.

[19]  G. Jack Lipovski,et al.  Parallel computing - theory and comparisons , 1987 .

[20]  Zarka Cvetanovic,et al.  The Effects of Problem Partitioning, Allocation, and Granularity on the Performance of Multiple-Processor Systems , 1987, IEEE Transactions on Computers.

[21]  Howard Jay Siegel,et al.  Using the multistage cube network topology in parallel supercomputers , 1989 .

[22]  Howard Jay Siegel.  The Theory Underlying the Partitioning of Permutation Networks , 1980, IEEE Transactions on Computers.