Automatic data layout transformation for heterogeneous many-core systems

Ying Yu Tseng, Yu Hao Huang, Bo-Cheng Lai, Jiun Liang Lin

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

5 Scopus citations

Abstract

Applying appropriate data structures is critical to attain superior performance in heterogeneous many-core systems. A heterogeneous many-core system is comprised of a host for control flow management, and a device for massive parallel data processing. However, the host and device require different types of data structures. The host prefers Array-of-Structures (AoS) to ease the programming, while the device requires Structure-of-Arrays (SoA) for efficient data accesses. The conflicted preferences cost excessive effort for programmers to transform the data structures between two parts. The separately designed kernels with different coding styles also cause difficulty in maintaining programs. This paper addresses this issue by proposing a fully automated data layout transformation framework. Programmers can maintain the code in AoS style on the host, while the data layout is converted into SoA when being transferred to the device. The proposed framework streamlines the design flow and demonstrates up to 177% performance improvement.

Original languageEnglish
Title of host publicationNetwork and Parallel Computing - 11th IFIP WG 10.3 International Conference, NPC 2014, Proceedings
PublisherSpringer Verlag
Pages208-219
Number of pages12
ISBN (Print)9783662449165
DOIs
StatePublished - 1 Jan 2014
Event11th IFIP WG 10.3 International Conference on Network and Parallel Computing, NPC 2014 - Ilan, Taiwan
Duration: 18 Sep 201420 Sep 2014

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8707 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference11th IFIP WG 10.3 International Conference on Network and Parallel Computing, NPC 2014
CountryTaiwan
CityIlan
Period18/09/1420/09/14

Keywords

  • GPGPU
  • data layout transformation
  • heterogeneous systems
  • many-core

Fingerprint Dive into the research topics of 'Automatic data layout transformation for heterogeneous many-core systems'. Together they form a unique fingerprint.

  • Cite this

    Tseng, Y. Y., Huang, Y. H., Lai, B-C., & Lin, J. L. (2014). Automatic data layout transformation for heterogeneous many-core systems. In Network and Parallel Computing - 11th IFIP WG 10.3 International Conference, NPC 2014, Proceedings (pp. 208-219). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8707 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-662-44917-2_18