Dataflow Execution of Hierarchically Tiled Arrays

Chih Chieh Yang, Juan C. Pichel, David A. Padua

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

As the parallelism in high-performance supercomputers continues to grow, new programming models become necessary to maintain programmer productivity at today’s levels. Dataflow is a promising execution model because it can represent parallelism at different granularity levels and to dynamically adapt for efficient execution. The downside is the low-level programming interface inherent to dataflow. We present a strategy to translate programs written in Hierarchically Tiled Arrays (HTA) to the dataflow API of Open Community Runtime (OCR) system. The goal is to enable program development in a convenient notation and at the same time take advantage of the benefits of a dataflow runtime system. Using HTA produces more comprehensive codes than those written using the dataflow runtime programming interface. Moreover, the experiments show that, for applications with high asynchrony and sparse data dependences, our implementation delivers superior performance than OpenMP using parallel for loops.

Original languageEnglish (US)
Title of host publicationEuro-Par 2019
Subtitle of host publicationParallel Processing - 25th International Conference on Parallel and Distributed Computing, Proceedings
EditorsRamin Yahyapour
PublisherSpringer
Pages304-316
Number of pages13
ISBN (Print)9783030293994
DOIs
StatePublished - 2019
Event25th International European Conference on Parallel and Distributed Computing, Euro-Par 2019 - Göttingen, Germany
Duration: Aug 26 2019Aug 30 2019

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11725 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference25th International European Conference on Parallel and Distributed Computing, Euro-Par 2019
CountryGermany
CityGöttingen
Period8/26/198/30/19

Keywords

  • Dataflow
  • High-level programming abstraction
  • Parallel algorithm
  • Parallel programming

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Dataflow Execution of Hierarchically Tiled Arrays'. Together they form a unique fingerprint.

Cite this