EnGPar

EnGPar, N-Graph Partitioning, scalable partitioning methods provide well balanced computational work with low communication costs. These methods improve the performance of finite element and finite volume based applications with element and vertex based unstructured mesh partitions, respectively, and particle-in-cell based applications that require dynamic control of particle distribution. EnGPar provides C/C++ and FORTRAN interfaces similar to those used for graph based partitioning tools. MPI is used for inter-processor communications.

CONTACT

Gerrett Diamond
Rensselaer Polytechnic Institute
518-276-6795
Send email


Cameron W. Smith
Rensselaer Polytechnic Institute
518-276-3401

Send email