Reduction of the effects of the communication delays in scientific algorithms on message passing MIMD architectures

Journal Article · Thu Jan 01 04:00:00 EST 1987 · SIAM J. Sci. Stat. Comput.; (United States)

DOI:https://doi.org/10.1137/0908016· OSTI ID:6566083

Saltz, J H; Naik, V K; Nicol, D M

The efficient implementation of algorithms on multiprocessor machines requires that the effects of communication delays be minimized. The effects of these delays on the performance of a model problem on a hypercube multiprocessor architecture is investigated, and methods are developed for increasing algorithm efficiency. This paper identifies methods for reducing communication traffic and overhead on a multiprocessor and reports the results of testing these methods on the Intel iPSC Hypercube. The authors examine methods for partitioning a problem's domain across processors, for reducing communication traffic during a global convergence check, for reducing the number of global convergence checks employed during an iteration, and for concurrently iterating on multiple time-steps in a time dependent problem. The empirical results show that use of these methods can markedly reduce a numerical problem's execution time.

• Research Organization: Institute for Computer Applications in Science and Engineering, NASA, Langley Research Center, Hampton, VA 23665
• OSTI ID: 6566083
• Journal Information: SIAM J. Sci. Stat. Comput.; (United States), Journal Name: SIAM J. Sci. Stat. Comput.; (United States) Vol. 8:1; ISSN SIJCD
• Country of Publication: United States
• Language: English

Related Subjects

99 GENERAL AND MISCELLANEOUS
990210* -- Supercomputers-- (1987-1989)
ALGORITHMS
ARRAY PROCESSORS
COMMUNICATIONS
COMPUTER ARCHITECTURE
COMPUTERS
DATA TRANSMISSION
DIGITAL COMPUTERS
EFFICIENCY
EQUIPMENT INTERFACES
GLOBAL ANALYSIS
ITERATIVE METHODS
MATHEMATICAL LOGIC
MATHEMATICS
PARALLEL PROCESSING
PERFORMANCE
PERFORMANCE TESTING
PROGRAMMING
SUPERCOMPUTERS
TESTING
TIME DEPENDENCE