Cooperative heterogeneous computing for parallel processing on CPU/GPU hybrids

Changmin Lee, Won W. Ro, Jean Luc Gaudiot

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

16 Citations (Scopus)

Abstract

This paper presents a cooperative heterogeneous computing framework which enables the efficient utilization of available computing resources of host CPU cores for CUDA kernels, which are designed to run only on GPU. The proposed system exploits at runtime the coarse-grain thread-level parallelism across CPU and GPU, without any source recompilation. To this end, three features including a work distribution module, a transparent memory space, and a global scheduling queue are described in this paper. With a completely automatic runtime workload distribution, the proposed framework achieves speedups as high as 3.08 compared to the baseline GPU-only processing.

Original languageEnglish
Title of host publicationProceedings - 16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012
Pages33-40
Number of pages8
DOIs
Publication statusPublished - 2012
Event16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012 - New Orleans, LA, United States
Duration: 2012 Feb 252012 Feb 25

Publication series

NameProceedings - Annual Workshop on Interaction between Compilers and Computer Architectures, INTERACT
ISSN (Print)1550-6207

Other

Other16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012
Country/TerritoryUnited States
CityNew Orleans, LA
Period12/2/2512/2/25

All Science Journal Classification (ASJC) codes

  • General Engineering

Fingerprint

Dive into the research topics of 'Cooperative heterogeneous computing for parallel processing on CPU/GPU hybrids'. Together they form a unique fingerprint.

Cite this