Dynamic detection of uniform and affine vectors in GPGPU computations
ELIAUS, Universite de Perpignan, 66860 Perpignan, France
EURO-PAR 2009 – Parallel Processing Workshops, Lecture Notes in Computer Science, 2010, Volume 6043/2010, 46-55
@conference{collange2010dynamic,
title={Dynamic detection of uniform and affine vectors in GPGPU computations},
author={Collange, S. and Defour, D. and Zhang, Y.},
booktitle={Euro-Par 2009–Parallel Processing Workshops},
pages={46–55},
year={2010},
organization={Springer}
}
We present a hardware mechanism which dynamically detects uniform and affine vectors used in SPMD architecture such as Graphics Processing Units, to minimize pressure on the register file and reduce power consumption with minimal architectural modifications. A preliminary experimental analysis conducted with the Barra simulator shows that this optimization can benefit up to 34 % of register file reads and 22 % of the computations in common GPGPU applications.
April 19, 2011 by hgpu
