Abstract
A key problem for parallelizing compilers is to find the good tradeoff between memory expansion and parallelism. This paper is a new step towards solving this problem. A framework for parallel execution order and storage mapping computation is designed, allowing time and space optimization. Constrained expansion a theoretical model for expansion strategies--is shown to be very useful in this context.
Cite
CITATION STYLE
Cohen, A. (1999). Parallelization via constrained storage mapping optimization. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1615, pp. 83–94). Springer Verlag. https://doi.org/10.1007/BFb0094913
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
