By Topic

Extending OpenMP for heterogeneous chip multiprocessors

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

2 Author(s)
Liu, F. ; Inst. for Sci. Comput., Wayne State Univ., Detroit, MI, USA ; Chaudhary, V.

The emergence of system-on-chip (SOC) design shows the growing popularity of the integration of multiple-processors into one chip. We propose that high-level abstraction of parallel programming like OpenMP is suitable for chip multiprocessors. For SOCs, the heterogeneity exists within one chip such that it may have different types of multiprocessors, e.g. RISC-like processors or DSP-like processors. Incorporating different processors into OpenMP is challenging. We present our solutions to extend OpenMP directives to tackle this heterogeneity. Several optimization techniques are proposed to utilize advanced architecture features of our target SOC, the software scalable system on chip (3SoC). Preliminary performance evaluation shows scalable speedup using different types of processors and performance improvement through individual optimization.

Published in:

Parallel Processing, 2003. Proceedings. 2003 International Conference on

Date of Conference:

6-9 Oct. 2003