Skip to Main content Skip to Navigation
Book sections

UMPa: A Multi-objective, multi-level partitioner for communication minimization

Abstract : We propose a directed hypergraph model and a refinement heuristic to distribute communicating tasks among the processing units in a distributed memory setting. The aim is to achieve load balance and minimize the maximum data sent by a processing unit. We also take two other communication metrics into account with a tie-breaking scheme. With this approach, task distributions causing an excessive use of network or a bottleneck processor which participates to almost all of the communication are avoided. We show on a large number of problem instances that our model improves the maximum data sent by a processor up to 34% for parallel environments with 4, 16, 64 and 256 processing units compared to the state of the art which only minimizes the total communication volume.
Complete list of metadata

Cited literature [20 references]  Display  Hide  Download

https://hal.inria.fr/hal-00763563
Contributor : Equipe Roma <>
Submitted on : Thursday, December 19, 2019 - 5:01:36 PM
Last modification on : Thursday, December 19, 2019 - 5:24:10 PM
Long-term archiving on: : Friday, March 20, 2020 - 10:57:51 PM

File

umpa-dimacs.pdf
Files produced by the author(s)

Identifiers

Collections

Citation

Umit Catalyurek, Mehmet Deveci, Kamer Kaya, Bora Uçar. UMPa: A Multi-objective, multi-level partitioner for communication minimization. David A. Bader; Henning Meyerhenke; Peters Sanders; Dorothea Wagner. Graph Partitioning and Graph Clustering 2012, 588, AMS, pp.53--66, 2013, Contemporary Mathematics, ⟨10.1090/conm/588/11704⟩. ⟨hal-00763563⟩

Share

Metrics

Record views

411

Files downloads

193