Work

Automatic data/program partitioning using the single assignment principle

Year: 1989

Type: conference-paper

Abstract: Loosely-coupled MIMD architectures do not suffer from memory contention; hence large numbers of processors may be utilized. The main problem, however, is how to partition data and programs in order to exploit the available parallelism. In this paper we show that efficient schemes for automatic data/program partitioning and synchronization may be employed if single assignment is used. Using simulations of program loops common to scientific computations (the Livermore Loops), we demonstrate that only a small fraction of data accesses are remote and thus the degradation in network performance due to multi-processing is minimal.

Source: ACM/IEEE Conference on Supercomputing (SC)

DOI: 10.1145/76263.76324

Authors: L. Bic, M. D. Nagel, J. M. A. Roy

Institution: University of California, Irvine

Language: English

Topic: Parallel Computing and Optimization Techniques

Subfield: Hardware and Architecture

Field: Computer Science

Domain: Physical Sciences

Open Access status: bronze

Funder: National Science Foundation

Award: Efficient Execution of Dataflow Programs