Skip to Main content Skip to Navigation
Conference papers

Maximizing Communication Overlap with Dynamic Program Analysis

Emmanuelle Saillard 1 Koushik Sen 2 Wim Lavrijsen 2 Costin Iancu 2 
1 STORM - STatic Optimizations, Runtime Methods
LaBRI - Laboratoire Bordelais de Recherche en Informatique, Inria Bordeaux - Sud-Ouest
Abstract : We present a dynamic program analysis approach to optimize communication overlap in scientific applications. Our tool instruments the code to generate a trace of the application's memory and synchronization behavior. An offline analysis determines the program optimal points for maximal overlap when considering several programming constructs: nonblocking one-sided communication operations, non-blocking collectives and bespoke synchronization patterns and operations. Feedback about possible transformations is presented to the user and the tool can perform the directed transformations, which are supported by a lightweight runtime. The value of our approach comes from: 1) the ability to optimize across boundaries of software modules or libraries, while specializing for the intrinsics of the underlying communication runtime; and 2) providing upper bounds on the expected performance improvements after communication optimizations. We have reduced the time spent in communication by as much as 64% for several applications that were already aggressively optimized for overlap; this indicates that manual optimizations leave untapped performance. Although demonstrated mainly for the UPC programming language, the methodology can be easily adapted to any other communication and synchronization API.
Complete list of metadata

Cited literature [38 references]  Display  Hide  Download
Contributor : Emmanuelle Saillard Connect in order to contact the contributor
Submitted on : Wednesday, November 28, 2018 - 10:18:02 AM
Last modification on : Saturday, June 25, 2022 - 10:38:44 AM


Files produced by the author(s)


  • HAL Id : hal-01937407, version 1



Emmanuelle Saillard, Koushik Sen, Wim Lavrijsen, Costin Iancu. Maximizing Communication Overlap with Dynamic Program Analysis. International Conference on High Performance Computing in Asia-Pacific Region, Jan 2018, Tokyo, Japan. ⟨hal-01937407⟩



Record views


Files downloads