Scalability challenges in current MPI one-sided implementations

Xin Zhao, Pavan Balaji, William D Gropp

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

MPI one-sided or remote memory access (RMA) communication provides a different execution model from traditional two-sided or group communication and is better suited for some classes of applications. However, current implementations of MPI RMA are notorious for their inability to scale to large systems or problem sizes. In this paper, we present a study of the RMA infrastructure in popular open-source MPI implementations. Our objective is to identify critical scalability limitations with respect to memory usage in these implementations. We then perform a thorough evaluation on two cluster computers to demonstrate those scalability limitations, and we provide suggestions on how they can be alleviated.

Original languageEnglish (US)
Title of host publicationProceedings - 15th International Symposium on Parallel and Distributed Computing, ISPDC 2016
EditorsRiqing Chen, Dan Grigoras, Chunming Rong
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages38-47
Number of pages10
ISBN (Electronic)9781509041527
DOIs
StatePublished - Apr 18 2017
Event15th International Symposium on Parallel and Distributed Computing, ISPDC 2016 - Fuzhou, Fujian, China
Duration: Jul 8 2016Jul 10 2016

Publication series

NameProceedings - 15th International Symposium on Parallel and Distributed Computing, ISPDC 2016

Other

Other15th International Symposium on Parallel and Distributed Computing, ISPDC 2016
CountryChina
CityFuzhou, Fujian
Period7/8/167/10/16

    Fingerprint

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Hardware and Architecture
  • Information Systems

Cite this

Zhao, X., Balaji, P., & Gropp, W. D. (2017). Scalability challenges in current MPI one-sided implementations. In R. Chen, D. Grigoras, & C. Rong (Eds.), Proceedings - 15th International Symposium on Parallel and Distributed Computing, ISPDC 2016 (pp. 38-47). [7904267] (Proceedings - 15th International Symposium on Parallel and Distributed Computing, ISPDC 2016). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ISPDC.2016.14