Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
skip to main content
10.1145/3332186.3333042acmotherconferencesArticle/Chapter ViewAbstractPublication PagespearcConference Proceedingsconference-collections
research-article

The PreSync Project: Synchronization Automation in the Cactus Framework

Published: 28 July 2019 Publication History

Abstract

The Einstein Toolkit (ETK) is an open-source software platform for computational simulations in relativistic astrophysics and gravitational physics. The Cactus Framework (Cactus) comprises the core component of the ETK and handles creation of distributed data structures, parallelism, I/O, checkpointing, etc. The current Cactus scheduling system relies on the manual synchronization of ghost zones for grid functions (i.e. distributed matrices). Unfortunately, deciding which subroutines should synchronize which grid functions is a non-trivial problem. Incorrect synchronization can result in over-synchronization (a performance problem) or under-synchronization (an error). This issue also creates a barrier for new users and collaborators. We have developed a new scheduling and synchronization method for Cactus to improve the accessibility and efficiency of the ETK. Synchronization and application of boundary conditions are now handled automatically by Cactus, and performance has been improved by as much as twenty percent. The new approach requires each subroutine to have "read" and "write" declarations for individual grid functions. Before scheduled subroutines run, Cactus checks these declarations and performs any synchronization or boundary condition updates as needed. This method removes the difficulty of deciding where synchronization should take place and removes unneeded synchronizations. The only knowledge required to provide these declarations is what grid functions a given subroutine uses and on which parts of the grid they are read or written. In addition, these changes have reduced the number of synchronizations by approximately five percent in unigrid simulations. This new system is a first step in improving Cactus' scheduling methods to use more advanced and scalable techniques.

References

[1]
James Bordner and Michael L Norman. 2012. Enzo-P/Cello: Scalable adaptive mesh refinement for astrophysics and cosmology. In Proceedings of the Extreme Scaling Workshop. University of Illinois at Urbana-Champaign, 4.
[2]
Greg L Bryan, Michael L Norman, Brian W O'Shea, Tom Abel, John H Wise, Matthew J Turk, Daniel R Reynolds, David C Collins, Peng Wang, Samuel W Skillman, et al. 2014. Enzo: An adaptive mesh refinement code for astrophysics. The Astrophysical Journal Supplement Series 211, 2 (2014), 19.
[3]
Tom Goodale, Gabrielle Allen, Gerd Lanfermann, Joan Massó, Thomas Radke, Edward Seidel, and John Shalf. 2002. The Cactus Framework and Toolkit: Design and Applications. In High Performance Computing for Computational Science --- VECPAR 2002, José M.L.M. Palma, A. Augusta Sousa, Jack Dongarra, and Vicente Hernández (Eds.). Lecture Notes in Computer Science, Vol. 2565. Springer-Verlag, Berlin, 197--227.
[4]
Hartmut Kaiser, Thomas Heller, Bryce Adelstein-Lelbach, Adrian Serio, and Dietmar Fey. 2014. Hpx: A task based programming model in a global address space. In Proceedings of the 8th International Conference on Partitioned Global Address Space Programming Models. ACM, 6.
[5]
Laxmikant V Kale and Sanjeev Krishnan. 1993. CHARM++: a portable concurrent object oriented system based on C++. Vol. 28. Citeseer.
[6]
Frank Löffler, Joshua Faber, Eloisa Bentivegna, Tanja Bode, Peter Diener, Roland Haas, Ian Hinder, Bruno C. Mundim, Christian D. Ott, Erik Schnetter, Gabrielle Allen, Manuela Campanelli, and Pablo Laguna. 2012. The Einstein Toolkit: A Community Computational Infrastructure for Relativistic Astrophysics. Classical and Quantum Gravity 29, 11 (May 2012).
[7]
Dominic C Marcello, Kundan Kadam, Geoffrey C Clayton, Juhan Frank, Hartmut Kaiser, and Patrick M Motl. 2016. Introducing Octo-tiger/HPX: Simulating interacting binaries with adaptive mesh refinement and the fast multipole method. Proceedings of Science (2016), 13--17.

Recommendations

Comments

Information & Contributors

Information

Published In

cover image ACM Other conferences
PEARC '19: Practice and Experience in Advanced Research Computing 2019: Rise of the Machines (learning)
July 2019
775 pages
ISBN:9781450372275
DOI:10.1145/3332186
  • General Chair:
  • Tom Furlani
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].

Publisher

Association for Computing Machinery

New York, NY, United States

Publication History

Published: 28 July 2019

Permissions

Request permissions for this article.

Check for updates

Author Tags

  1. ACM proceedings
  2. Cactus Framework
  3. Einstein Toolkit
  4. frameworks
  5. ghost zones
  6. synchronization

Qualifiers

  • Research-article
  • Research
  • Refereed limited

Conference

PEARC '19

Acceptance Rates

Overall Acceptance Rate 133 of 202 submissions, 66%

Contributors

Other Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

  • 0
    Total Citations
  • 56
    Total Downloads
  • Downloads (Last 12 months)3
  • Downloads (Last 6 weeks)0
Reflects downloads up to 10 Feb 2025

Other Metrics

Citations

View Options

Login options

View options

PDF

View or Download as a PDF file.

PDF

eReader

View online with eReader.

eReader

Figures

Tables

Media

Share

Share

Share this Publication link

Share on social media