Communication and Topology-aware Load Balancing in Charm++ with TreeMatch
| Autor: | Guillaume Mercier, Esteban Meneses, Emmanuel Jeannot, Francois Tessier, Gengbin Zheng |
|---|---|
| Přispěvatelé: | Laboratoire Bordelais de Recherche en Informatique (LaBRI), Université de Bordeaux (UB)-Centre National de la Recherche Scientifique (CNRS)-École Nationale Supérieure d'Électronique, Informatique et Radiocommunications de Bordeaux (ENSEIRB), Efficient runtime systems for parallel architectures (RUNTIME), Inria Bordeaux - Sud-Ouest, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Bordeaux (UB)-Centre National de la Recherche Scientifique (CNRS), Department of Computer Science [UIUC] (UIUC), University of Illinois at Urbana-Champaign [Urbana], University of Illinois System-University of Illinois System, plafrim, Joint-Laboratory for Petascale Computing, Université de Bordeaux (UB)-École Nationale Supérieure d'Électronique, Informatique et Radiocommunications de Bordeaux (ENSEIRB)-Centre National de la Recherche Scientifique (CNRS) |
| Jazyk: | angličtina |
| Rok vydání: | 2013 |
| Předmět: |
020203 distributed computing
Multi-core processor Computer science Distributed computing message passing Message passing load balancing 020207 software engineering 02 engineering and technology Load balancing (computing) Topology process placement 0202 electrical engineering electronic engineering information engineering Resource allocation communication cost [INFO.INFO-DC]Computer Science [cs]/Distributed Parallel and Cluster Computing [cs.DC] |
| Zdroj: | IEEE Cluster 2013 IEEE Cluster 2013, Sep 2013, Indianapolis, United States CLUSTER |
| Popis: | International audience; Programming multicore or manycore architectures is a hard challenge particularly if one wants to fully take advantage of their computing power. Moreover, a hierarchical topology implies that communication performance is heterogeneous and this characteristic should also be exploited. We developed two load balancers for Charm++ that take into account both aspects, depending on the fact that the application is compute-bound or communication-bound. This work is based on our TREEMATCH library that computes process placement in order to reduce an application communication costs based on the hardware topology. We show that the proposed load-balancing schemes manage to improve the execution times for the two aforementioned classes of parallel applications. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|