Parallel graph partitioning algorithm for a message-passing multiprocessor

Journal Article · Tue Dec 01 00:00:00 EST 1987 · Int. J. Parallel Program.; (United States)

DOI:https://doi.org/10.1007/BF01388998· OSTI ID:6264875

Gilbert, J R; Zmijewski, E

The authors develop a parallel algorithm for partitioning the vertices of a graph into p greater than or equal to 2 sets in such a way that few edges connect vertices in different sets. The algorithm is intended for a message-passing multiprocessor system, such as the hypercube, and is based on the Kernighan-Lin algorithm for finding small edge separators on a single processor. They use this parallel partitioning algorithm to find orderings for factoring large sparse symmetric positive definite matrices. These orderings not only reduce fill, but also result in good processor utilization and low communication overheat during the factorization. They provide a complexity analysis of the algorithm, as well as some numerical results from an Intel hypercube and a hypercube simulator.

• Research Organization: Cornell Univ., Ithaca, NY (United States)
• OSTI ID: 6264875
• Journal Information: Int. J. Parallel Program.; (United States), Vol. 16:6
• Country of Publication: United States
• Language: English

Related Subjects

99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE
ARRAY PROCESSORS
PARALLEL PROCESSING
COMPUTER GRAPHICS
ALGORITHMS
COMPUTERIZED SIMULATION
DATA TRANSMISSION
EFFICIENCY
FACTORIZATION
MATRICES
SYSTEMS ANALYSIS
COMMUNICATIONS
MATHEMATICAL LOGIC
PROGRAMMING
SIMULATION
990210* - Supercomputers- (1987-1989)