Title
Comparing Nested Dissection Orderings for Parallel Sparse Matrix Factorization
Abstract
In this paper we compare nested dissection orderings obtained bydifferent graph bisection heuristics. In the context of parallel sparsematrix factorization the quality of an ordering is not only determinedby its fill reducing capability, but also depends on the difficultywith which a balanced mapping of the load onto the processorsof the parallel computer can be found. Our analysis showsthat sophisticated local bisection heuristics combined with the multilevelmethod result in high ...
Year
Venue
Keywords
1995
PDPTA
sparse matrix,parallel computer
Field
DocType
Citations 
Sparse matrix factorization,Computer science,Parallel computing,Sparse approximation,Algorithm,Nested dissection,Incomplete LU factorization,Band matrix,Sparse matrix
Conference
2
PageRank 
References 
Authors
0.37
3
3
Name
Order
Citations
PageRank
Jürgen Schulze120.71
Ralf Diekmann228128.99
Robert Preis333625.95