A uGNI-Based MPICH2 Nemesis Network Module for the Cray XE

  • Howard Pritchard
  • Igor Gorodetsky
  • Darius Buntinas
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6960)


Recent versions of MPICH2 have featured Nemesis – a scalable, high-performance, multi-network communication subsystem. Nemesis provides a framework for developing Network Modules (Netmods) for interfacing the Nemesis subsystem to various high speed network protocols. Cray has developed a user-level Generic Network Interface (uGNI) for interfacing MPI implementations to the internal high speed network of Cray XE and follow-on computer systems. This paper describes the design of a uGNI Netmod for the MPICH2 nemesis subsystem. MPICH2 performance data on the Cray XE are presented.


Network Module Message Length Defense Advance Research Project Agency Large Message Application Message 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Alverson, R., Roweth, D., Kaplan, L.: The Gemini System Interconnect. In: Symposium on High-Performance Interconnects, vol. 0, pp. 83–87 (2010)Google Scholar
  2. 2.
    Buntinas, D., Mercier, G., Gropp, W.: Design and Evaluation of Nemesis, a Scalable, Low-Latency, Message-Passing Communication Subsystem. In: CCGRID 2006, pp. 521–530 (2006)Google Scholar
  3. 3.
    Cray, Inc.: Cray Software Document S-2446-3103: Using the GNI and DMAPP APIs (March 2011)Google Scholar
  4. 4.
    Fault Tolerance Working Group: Run-though Stabilization Interfaces and Semantics,
  5. 5.
    Lai, P., Balaji, P., Thakur, R., Panda, D.K.: ProOnE: a General-purpose Protocol Onload Engine for Multi- and Many-core Architectures. Computer Science - R&D, 133–142 (2009)Google Scholar
  6. 6.
  7. 7.
    MPICH2–Nemesis: Nemesis Network Module API,
  8. 8.
    Network–Based Computing Laboratory: MVAPICH: MPI over Infiniband, 10GigE/iWARP and RoCE,
  9. 9.
    Wyckoff, P., Wu, J.: Memory Registration Caching Correctness. In: Proceedings of CCGrid 2005. IEEE Computer Society, Los Alamitos (2005)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  • Howard Pritchard
    • 1
  • Igor Gorodetsky
    • 1
  • Darius Buntinas
    • 2
  1. 1.Cray Inc.USA
  2. 2.Argonne National LaboratoryUSA

Personalised recommendations