A Proposal for User-defined Reductions in OpenMP

Conference · Mon Mar 22 00:00:00 EDT 2010

DOI:https://doi.org/10.1007/978-3-642-13217-9_4· OSTI ID:1020356

Duran, A; Ferrer, R; Klemm, M; de Supinski, B R; Ayguade, E

Reductions are commonly used in parallel programs to produce a global result from partial results computed in parallel. Currently, OpenMP only supports reductions for primitive data types and a limited set of base language operators. This is a significant limitation for those applications that employ user-defined data types (e. g., objects). Implementing manual reduction algorithms makes software development more complex and error-prone. Additionally, an OpenMP runtime system cannot optimize a manual reduction algorithm in ways typically applied to reductions on primitive types. In this paper, we propose new mechanisms to allow the use of most pre-existing binary functions on user-defined data types as User-Defined Reduction (UDR) operators. Our measurements show that our UDR prototype implementation provides consistently good performance across a range of thread counts without increasing general runtime overheads.

• Research Organization: Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)
• Sponsoring Organization: USDOE
• DOE Contract Number: W-7405-ENG-48
• OSTI ID: 1020356
• Report Number(s): LLNL-CONF-426239; TRN: US201116%%285
• Resource Relation: Journal Volume: 6132; Conference: Presented at: 6th International Workshop on OpenMP, Tsukuba, Japan, Jun 14 - Jun 16, 2010
• Country of Publication: United States
• Language: English

Related Subjects

97 MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE
ALGORITHMS
IMPLEMENTATION
PERFORMANCE