Extending a highly parallel data mining algorithm to the intel ® many integrated core architecture

Alexander Heinecke, Michael Klemm, Dirk Pflüger, Arndt Bode, Hans Joachim Bungartz

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

20 Scopus citations

Abstract

Extracting knowledge from vast datasets is a major challenge in data-driven applications, such as classification and regression, which are mostly compute bound. In this paper, we extend our SG + + algorithm to the Intel® Many Integrated Core Architecture (Intel® MIC Architecture). The ease of porting an application to Intel MIC Architecture is shown: porting existing SSE code is very easy and straightforward. We evaluate the current prototype pre-release coprocessor board codenamed Intel® "Knights Ferry". We utilize the pragma-based offloading programming model offered by the Intel® Composer XE for Intel MIC Architecture, generating both the host and the coprocessor code. We compare the achieved performance with an NVIDIA C2050 accelerator and show that the pre-release Knights Ferry coprocessor delivers better performance than the C2050 and exceeds the C2050 when comparing the productivity aspect of implementing algorithms for the coprocessors.

Original languageEnglish
Title of host publicationEuro-Par 2011
Subtitle of host publicationParallel Processing Workshops - CCPI, CGWS, HeteroPar, HiBB, HPCVirt, HPPC, HPSS, MDGS, ProPer, Resilience, UCHPC, VHPC, Revised Selected Papers
PublisherSpringer Verlag
Pages375-384
Number of pages10
EditionPART 2
ISBN (Print)9783642297397
DOIs
StatePublished - 2012
Externally publishedYes
Event17th Parallel Processing Workshops, Euro-Par 2011: CCPI 2011, CGWS 2011, HeteroPar 2011, HiBB 2011, HPCVirt 2011, HPPC 2011, HPSS 2011, MDGS 2011, ProPer 2011, Resilience 2011, UCHPC 2011, VHPC 2011 - Bordeaux, France
Duration: 29 Aug 20112 Sep 2011

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 2
Volume7156 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference17th Parallel Processing Workshops, Euro-Par 2011: CCPI 2011, CGWS 2011, HeteroPar 2011, HiBB 2011, HPCVirt 2011, HPPC 2011, HPSS 2011, MDGS 2011, ProPer 2011, Resilience 2011, UCHPC 2011, VHPC 2011
Country/TerritoryFrance
CityBordeaux
Period29/08/112/09/11

Keywords

  • GPGPU
  • NVIDIA Fermi
  • accelerators
  • coprocessors
  • data mining
  • sparse grids

Fingerprint

Dive into the research topics of 'Extending a highly parallel data mining algorithm to the intel ® many integrated core architecture'. Together they form a unique fingerprint.

Cite this