Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

CLOMP: Accurately Characterizing OpenMP Application Overheads

Conference ·
Despite its ease of use, OpenMP has failed to gain widespread use on large scale systems, largely due to its failure to deliver sufficient performance. Our experience indicates that the cost of initiating OpenMP regions is simply too high for the desired OpenMP usage scenario of many applications. In this paper, we introduce CLOMP, a new benchmark to characterize this aspect of OpenMP implementations accurately. CLOMP complements the existing EPCC benchmark suite to provide simple, easy to understand measurements of OpenMP overheads in the context of application usage scenarios. Our results for several OpenMP implementations demonstrate that CLOMP identifies the amount of work required to compensate for the overheads observed with EPCC. Further, we show that CLOMP also captures limitations for OpenMP parallelization on NUMA systems.
Research Organization:
Lawrence Livermore National Laboratory (LLNL), Livermore, CA
Sponsoring Organization:
USDOE
DOE Contract Number:
W-7405-ENG-48
OSTI ID:
929528
Report Number(s):
LLNL-CONF-401267
Country of Publication:
United States
Language:
English

References (6)

Flash code: studying astrophysical thermonuclear flashes journal March 2000
NAMD: Biomolecular Simulation on Thousands of Processors conference January 2002
Improving the computational intensity of unstructured mesh applications conference January 2005
The Community Climate System Model Version 3 (CCSM3) journal June 2006
ScaLAPACK Users' Guide book January 1997
Efficient Management of Parallelism in Object-Oriented Numerical Software Libraries book January 1997

Similar Records

CLOMP: Accurately Characterizing OpenMP Application Overheads
Journal Article · 2008 · International Journal of Parallel Programming · OSTI ID:956854

Clomp
Software · 2007 · OSTI ID:1257767

Analysis of OpenMP 4.5 Offloading in Implementations: Correctness and Overhead
Journal Article · 2019 · Parallel Computing · OSTI ID:1648853