Predictive performance tuning of open ACC accelerated applications

Shahzeb Siddiqui, Saber Feki

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

3 Scopus citations


GPUs are gradually becoming mainstream in supercomputing as their capabilities to significantly accelerate a large spectrum of scientific applications have been identified and proven. Moreover, with the introduction of directive based programming models such as OpenACC, these devices are becoming more accessible and practical to use by a larger scientific community. However, performance optimization of OpenACC applications usually requires an indepth knowledge of the hardware and software specifications. We suggest a prediction-based performance tuning mechanism to quickly tune OpenACC parameters to dynamically adapt to the execution environment on a given system. This approach is applied to a finite difference kernel to tune the OpenACC gang and vector clauses for mapping the computations into the underlying accelerator architecture. Our experiments show a good performance improvement against the default compiler parameters and a faster tuning by an order of magnitude compared to the brute force search tuning.

Original languageEnglish (US)
Title of host publicationSupercomputing - 29th International Conference, ISC 2014, Proceedings
PublisherSpringer Verlag
Number of pages2
Volume8488 LNCS
ISBN (Print)9783319075174
StatePublished - 2014
Event29th International Supercomputing Conference, ISC 2014 - Leipzig, Germany
Duration: Jun 22 2014Jun 26 2014

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8488 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Other29th International Supercomputing Conference, ISC 2014

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of 'Predictive performance tuning of open ACC accelerated applications'. Together they form a unique fingerprint.

Cite this