We are proud to announce the release of a major update of our graph partitioning framework KaHiP (Karlsruhe High Quality Graph Partitioning). KaHiP is a family of high quality graph partitioning programs. It contains various graph partitioning algorithms that can be configured to either achieve the best known partitions for many standard benchmark instances or to be a good trade-off between partition quality and running time. Since the last major release we added:<p>- node ordering algorithms to compute fill-in reduced node orderings
- ILP based exact solvers and partition improvement algorithms
- global multisection process mapping algorithms
- and a lot of minor improvements of our system<p>This is round of improvements and extensions is due to
Alexandra Henzinger, Alexander Noe, Wolfgang Ost, Daniel Seemaier<p>The code is available under MIT Licence.<p>* open source implementation / website
<a href="https://kahip.github.io" rel="nofollow">https://kahip.github.io</a><p>* github
<a href="https://github.com/KaHIP" rel="nofollow">https://github.com/KaHIP</a><p>We are glad for any comments, stars and error reports (or even bug fixes) that you send us.<p>Alexandra Henzinger, Alexander Noe, Wolfgang Ost, Peter Sanders, Christian Schulz, Daniel Seemaier
Karlsruhe Institute of Technology (KIT)
University of Vienna