Graph partitioning models for parallel computing

497 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

On Two-Dimensional Sparse Matrix Partitioning: Models, Methods, and a Recipe (2010, methodology): “Graph and hypergraph partitioning models have been used for one-dimensional (1D) partitioning of sparse matrices [4, 5, 8, 9, 19, 20, 25, 26, 30, 32, 37].”
Sourcebook of parallel computing (2003, methodology): “In particular, on message-passing architectures, minimizing the maximum number of message start-ups that any one processor must perform can sometimes be more important than minimizing the communications volume [456].”
Multilevel algorithms for partitioning power-law graphs (2006, background): “These are the static graph partitioning, the graph repartitioning, and the multi-constraint, multi-objective graph-partitioning [14, 13, 28].”
Recent Advances in Graph Partitioning (2013, methodology): “For instance when GP is used in parallel computing to map the graph nodes to different processors, the communication volume is often more appropriate than the cut [100].”
Performance of fully coupled domain decomposition preconditioners for finite element transport/reaction simulations (2005, methodology): “[7,11,12, 13 ,28]). The choices also include the selection of subdomain solver such as an incomplete LU factorization (ILU) (with further options for dropping nonzeros in the factorizations and ordering equations within a subdomain [1] ), and the introduction of a coarse grid solve [10,37].”