Loop Parallelization by means of Affine Transformations
This project aims at building a set of tools that will allow the user to extract independent threads of computation (i.e., parallelism at a coarse grain) within loops, according tho the affine transformation framework.
The main contribution of this work will be the derivation of formal algorithms that will allow more parallelism to be extract than what is currently possible with state of the art tools.
Bibliography
- Anna Beletska. “Extracting coarse-grained parallelism with the Affine Transformation Framework and its limitations”. In Electronic Modelling, no.5, (2006)
- Anna Beletska, Pierluigi San Pietro. “Using the Affine Transformation Framework for Computer Simulation”. In Proceedings of International Conference on Advanced Computer Systems (ACS'2006)