The Method Of Parallel Recognition And Parallel Optimization Based On Data Dependence With Sparse Matrix

Abstract: for application programs in scientific and technological fields have grown increasingly large and complex, it is becoming more difficult to parallelize these programs by hand using message-passing libraries. To reduce this difficulty, we are researching the compilation technology for serial program automatic parallelization. In this paper, the author puts forward a kind of parallel recognition algorithm in parallelization compiler with sparse matrix to reduce memory consumption and time complexity. In the algorithm the author adopts the idea of the medium grain parallel.

REFERENCES

[1]. JIN Cheng-zhi. �The Construction Principles and Implementation Techniques of Compilers� , Higher Education Press , Beijing , 2000.

[2]. HU Yan-li, ZHANG Wei-ming. �A dynamic scheduling algorithm of parallel coarse grain tasks in computational grid based on time-balancing strategy�, .Journal of Chinese Computer System , 2008

[3]. SHEN Zhi-yu ,HUZi-ang ,�Methods of Parallel Compilation�, National Defence Industry Press, Beijing , 2001.

[4]. CHEN Guo-liang . �Design and Analysis of Parallel Algorithm �, Higher Education Press , Beijing , 2002 .

[5]. Stanford Compiler Group. �SUIF Compiler System Version1.0 ,US� . Stanford University, 1994.

[6]. LI Jing, ZANG Bin-yu , �Automatic Parallelism Detection for One Kind of Irregular Problems�, Journal of Software , 2002

[7]. M. Grieble, �Automatic Parallelization of Loop Programs for Distributed Memory Architectures�, FMI, University of Passau, 2004.

[8]. L. N. Pouchet, C. Bastoul, A. Cohen. �Iterative Optimization in the Polyhedral Model: part I, One Dimensional Time�, 2007.

[9]. Zhao Yan , Lei Liu , Li Ma . � The Method of Parallel Optimization and Parallel Recognition Based on Data Dependence � , 2009 IEEE.