Unchaining in design-space optimization of streaming applications

Shobana Padmanabhan, Yixin Chen, Roger D. Chamberlain

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

2 Scopus citations

Abstract

Data-streaming applications are frequently pipelined and deployed on hybrid systems to meet performance requirements and resource constraints. With freedom in the design of algorithms and architectures, the search complexity can explode. A popular approach to reducing search complexity is to decompose the search space while preserving optimality. We present a novel decomposition technique called unchaining that partitions the problem such that the resulting sub problems are less complex. Thanks to unchaining, the number of sub problems from the decomposition is linear in the number of chained blocks in the variable-constraint matrix (instead of being their product). Finally, we present a queueing network model and the quantitative search space reduction for a real world implementation of a bio sequence search application called BLASTN.

Original languageEnglish
Title of host publicationProceedings - 2013 3rd Workshop on Data-Flow Execution Models for Extreme Scale Computing, DFM 2013
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages63-70
Number of pages8
ISBN (Electronic)9781479952472
DOIs
StatePublished - Oct 8 2014
Event2013 3rd Workshop on Data-Flow Execution Models for Extreme Scale Computing, DFM 2013 - Edinburgh, United Kingdom
Duration: Sep 8 2013Sep 8 2013

Publication series

NameProceedings - 2013 3rd Workshop on Data-Flow Execution Models for Extreme Scale Computing, DFM 2013

Conference

Conference2013 3rd Workshop on Data-Flow Execution Models for Extreme Scale Computing, DFM 2013
Country/TerritoryUnited Kingdom
CityEdinburgh
Period09/8/1309/8/13

Keywords

  • decomposition of queueing networks
  • design-space exploration
  • domain-specific branch and bound

Fingerprint

Dive into the research topics of 'Unchaining in design-space optimization of streaming applications'. Together they form a unique fingerprint.

Cite this