Designing a High Performance OpenSHMEM Implementation Using Universal Common Communication Substrate as a Communication Middleware

  • Pavel Shamis
  • Manjunath Gorentla Venkata
  • Stephen Poole
  • Aaron Welch
  • Tony Curtis
Conference paper

DOI: 10.1007/978-3-319-05215-1_1

Part of the Lecture Notes in Computer Science book series (LNCS, volume 8356)
Cite this paper as:
Shamis P., Venkata M.G., Poole S., Welch A., Curtis T. (2014) Designing a High Performance OpenSHMEM Implementation Using Universal Common Communication Substrate as a Communication Middleware. In: Poole S., Hernandez O., Shamis P. (eds) OpenSHMEM and Related Technologies. Experiences, Implementations, and Tools. OpenSHMEM 2014. Lecture Notes in Computer Science, vol 8356. Springer, Cham

Abstract

OpenSHMEM is an effort to standardize the well-known SHMEM parallel programming library. The project aims to produce an open-source and portable SHMEM API and is led by ORNL and UH. In this paper, we optimize the current OpenSHMEM reference implementation, based on GASNet, to achieve higher performance characteristics. To achieve these desired performance characteristics, we have redesigned an important component of the OpenSHMEM implementation, the network layer, to leverage a low-level communication library designed for implementing parallel programming models called UCCS. In particular, UCCS provides an interface and semantics such as native atomic operations and remote memory operations to better support PGAS programming models, including OpenSHMEM. Through the use of microbenchmarks, we evaluate this new OpenSHMEM implementation on various network metrics, including the latency of point-to-point and collective operations. Furthermore, we compare the performance of our OpenSHMEM implementation with the state-of-the-art SGI SHMEM. Our results show that the atomic operations of our OpenSHMEM implementation outperform SGI’s SHMEM implementation by 3%. Its RMA operations outperform both SGI’s SHMEM and the original OpenSHMEM reference implementation by as much as 18% and 12% for gets, and as much as 83% and 53% for puts.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer International Publishing Switzerland 2014

Authors and Affiliations

  • Pavel Shamis
    • 1
  • Manjunath Gorentla Venkata
    • 1
  • Stephen Poole
    • 1
  • Aaron Welch
    • 2
  • Tony Curtis
    • 2
  1. 1.Extreme Scale Systems Center (ESSC)Oak Ridge National Laboratory (ORNL)USA
  2. 2.Computer Science DepartmentUniversity of Houston (UH)USA

Personalised recommendations