Skip to main content
OSTI.GOVU.S. Department of Energy Office of Scientific and Technical Information
U.S. Department of Energy
Office of Scientific and Technical Information
 

Multidimensional spectral load balancing

Patent ·
OSTI ID:870755
 [1];  [1]
  1. Albuquerque, NM
+ Show Author Affiliations

A method of and apparatus for graph partitioning involving the use of a plurality of eigenvectors of the Laplacian matrix of the graph of the problem for which load balancing is desired. The invention is particularly useful for optimizing parallel computer processing of a problem and for minimizing total pathway lengths of integrated circuits in the design stage.

Research Organization:
Sandia National Laboratories (SNL), Albuquerque, NM, and Livermore, CA
Assignee:
Sandia Corporation (Albuquerque, NM)
Patent Number(s):
US 5587922
Application Number:
08/680,718
OSTI ID:
870755
Country of Publication:
United States
Language:
English

References (7)

Multidimensional spectral load balancing report January 1993
An Efficient Heuristic Procedure for Partitioning Graphs journal February 1970
An improved spectral graph partitioning algorithm for mapping parallel computations report September 1992
Laplace eigenvalues of graphs—a survey journal November 1992
Partitioning of unstructured problems for parallel processing journal January 1991
An algorithm for quadrisection and its application to standard cell placement journal March 1988
Geometric algorithms for a minimum cost assignment problem conference January 1991

Similar Records

Multidimensional spectral load balancing
Technical Report · 1992 · OSTI ID:6691328
Multidimensional spectral load balancing
Technical Report · 1992 · OSTI ID:10135054
Method of up-front load balancing for local memory parallel processors
Patent · 1990 · OSTI ID:6597681

Related Subjects

/716/712/
apparatus
balancing
circuits
computer
computer process
design
desired
eigenvectors
graph
integrated
integrated circuit
integrated circuits
involving
laplacian
lengths
load
load balancing
matrix
method
minimizing
multidimensional
optimizing
parallel
parallel computer
particularly
particularly useful
partitioning
pathway
plurality
processing
spectral
stage
total
useful