OpenMP

Charm++ now comes with integrated runtime support for OpenMP. This allows Charm++ and AMPI users to stick OpenMP compiler directives inside their tasks to parallelize their work across the cores of a node. Our integrated runtime will cooperatively schedule work across the idle cores on a node, interleaving OpenMP task execution with that of other chares. The LLVM OpenMP runtime has been modified to use Charm++'s user-level threads and scheduler. It works with Clang, GCC, and ICC.

Papers / Talks

• 18-02

  2018
  Paper

  Multi-level Load Balancing with an Integrated Runtime Approach

  • Seonmyeong Bak
  • Harshitha Menon
  • Sam White
  • Matthias Diener
  • Laxmikant Vasudeo Kale
• 17-08

  2017
  Paper

  Integrating OpenMP into the Charm++ Programming Model

  • Seonmyeong Bak
  • Harshitha Menon
  • Sam White
  • Matthias Diener
  • Laxmikant Vasudeo Kale
• 16-19

  2016
  Paper

  Handling Transient and Persistent Imbalance Together in Distributed and Shared Memory

  • Harshitha Menon
  • Seonmyeong Bak
  • Phil Miller
  • Sam White
  • Nitin Bhat
  • Laxmikant Vasudeo Kale
• 11-04

  2011
  Paper

  Evaluation of Simple Causal Message Logging for Large-Scale Fault Tolerant HPC Systems

  • Esteban Meneses
  • Greg Bronevetsky
  • Laxmikant Vasudeo Kale
• 93-13

  1993
  Paper

  A Load Balancing Strategy For Prioritized Execution of Tasks

  • Amitabh Sinha
  • Laxmikant Vasudeo Kale