loading...
Maximum Loop Distribution and Fusion for Two-level Loops Considering Code Size
Las Vegas, Nevada, USA December 07-December 09
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/ISPAN.2005.588th International Symposium on Parall ...
 This Article 
 
PDF
HTML
 
 Share 
   
 Bibliographic References 
   
 Add to: 
 
Digg
Furl
Spurl
Blink
Simpy
Google
Del.icio.us
Y!MyWeb
 
 Search 
   
Meilin Liu, University of Texas at Dallas
Qingfeng Zhuge, University of Texas at Dallas
Zili Shao, Hong Kong Polytechnic University
Chun Xue, University of Texas at Dallas
Meikang Qiu, University of Texas at Dallas
Edwin H.-M. Sha, University of Texas at Dallas
In this paper, we propose a technique combining loop distribution with loop fusion to improve the timing performance without increasing the code size of the transformed loops. We first develop the loop distribution theorems that state the conditions distributing any two-level nested loop in the maximum way. Based on the loop distribution theorems, we design an algorithm to conduct maximum loop distribution. Then we propose a technique of maximum loop distribution with direct loop fusion, which performs maximum loop distribution followed by direct loop fusion. The experimental results show that the execution time of the transformed loops by our technique is reduced 41.9% on average compared to the original loops without the increase of the code size.
Citation:
Meilin Liu, Qingfeng Zhuge, Zili Shao, Chun Xue, Meikang Qiu, Edwin H.-M. Sha, "Maximum Loop Distribution and Fusion for Two-level Loops Considering Code Size," ispan, pp.126-131, 8th International Symposium on Parallel Architectures,Algorithms and Networks (ISPAN'05), 2005
Usage of this product signifies your acceptance of the Terms of Use.