DocumentCode :
1462992
Title :
Achieving full parallelism using multidimensional retiming
Author :
Passos, Nelson Luiz ; Sha, Edwin Hsing-Mean
Author_Institution :
Dept. of Comput. Sci., Midwestern State Univ., Wichita Falls, TX, USA
Volume :
7
Issue :
11
fYear :
1996
fDate :
11/1/1996 12:00:00 AM
Firstpage :
1150
Lastpage :
1163
Abstract :
Most scientific and digital signal processing (DSP) applications are recursive or iterative. Transformation techniques are usually applied to get optimal execution rates in parallel and/or pipeline systems. The retiming technique is a common and valuable transformation tool in one-dimensional problems, when loops are represented by data flow graphs (DFGs). In this paper, uniform nested loops are modeled as multidimensional data flow graphs (MDFGs). Full parallelism of the loop body, i.e., all nodes in the MDFG executed in parallel, substantially decreases the overall computation time. It is well known that, for one-dimensional DFGs, retiming can not always achieve full parallelism. Other existing optimization techniques for nested loops also can not always achieve full parallelism. This paper shows an important and counter-intuitive result, which proves that we can always obtain full-parallelism for MDFGs with more than one dimension. This result is obtained by transforming the MDFG into a new structure. The restructuring process is based on a multidimensional retiming technique. The theory and two algorithms to obtain full parallelism are presented in this paper. Examples of optimization of nested loops and digital signal processing designs are shown to demonstrate the effectiveness of the algorithms
Keywords :
data flow graphs; instruction sets; parallel architectures; signal processing; data flow graphs; digital signal processing applications; full parallelism; multidimensional data flow graphs; multidimensional retiming; optimal execution rates; optimization techniques; pipeline systems; restructuring process; retiming technique; uniform nested loops; Concurrent computing; Design optimization; Digital signal processing; Flow graphs; Multidimensional systems; Parallel processing; Pipelines; Process design; Signal design; Signal processing algorithms;
fLanguage :
English
Journal_Title :
Parallel and Distributed Systems, IEEE Transactions on
Publisher :
ieee
ISSN :
1045-9219
Type :
jour
DOI :
10.1109/71.544356
Filename :
544356
Link To Document :
بازگشت