Abstract
In this paper, we propose the sandglass-type parallelization technique for a doacross loop which has the characteristics of iterationbased parallelizing and software pipelining. We prove its effectiveness by comparing the sandglass-type to well-known three parallelization techniques: iteration-based, software pipelining, and a combination of doalltype parallel and sequential techniques. We conclude that the sandglasstype parallelization technique is the most effective among the techniques mentioned above in cases in which there are less than ten processing elements and the size of tasks with loop-carried dependences is smaller than the size of tasks lacking loop-carried dependence.
This is a preview of subscription content, log in via an institution.
Preview
Unable to display preview. Download preview PDF.
References
Cytron, R.: Doacross: Beyond Vectorization for Multiprocessors, Proc. of the Int. Conf. on Parallel Processing, (1986) 836–844
Kanamaru, T., Koseki, A., Komatsu, H. and Fukazawa, Y.: Loop Staging: A Loop Parallelization Technique for Shared Memory Multiprocessors, Proc. of Joint Symposium on Parallel Processing 1997, (1997) 197–204 in Japanese
Kodama, Y., Sakane, H., Sato, M., Yamana, H., Sakai, S. and Yamaguchi, Y.: The EM-X. Parallel Computer: Architecture and Basic Performance, Proc. 22nd Annual Int. Symp. on Computer Architecture, (1995) 14–23
Nakanishi, T., Joe, K., Polychronopoulos, C. D., Araki, K., and Fukuda, A.: Estimating Parallel Execution Time, of Loops with Loop-Carried Dependences, Proc. of Int. Conf. on Parallel Processing, Vol. 3, (1996) 61–69
Nichols, B., Buttlar, D. and Farrell, J. P.: Pthreads Programming, O'Reilly & Associates, Inc. (1996)
Padua, D. A., Kuck, D. J., and Lawrie, D. H.: High-Speed Multiprocessors and Compilation Techniques, IEEE Trans. on Comp., Vol. C-29, No. 9, (1980) 763–776
Takabatake, M., Honda, H., Osawa, N. and Yuba, T.: A Parallel Technique for a Single Doacross Loop, Proc. of Joint Symposium on Parallel Processing 1998, (1998) 367–374 in Japanese
Yamana, H., Tatebe, O., Koike, H., Kodama, Y., Sakane, H. and Yamaguchi, Y.: Loopacross: Beyond Doacross for Distributed Memory Multiprocessors, Proc. of the Second IASTED International Conference, (1998) 229–235
Zima, Z. and Chapman, B.: Supercompilers for Parallel and Vector Computer, Addison Wesley (1990)
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1999 Springer-Verlag
About this paper
Cite this paper
Takabatake, M., Honda, H., Yuba, T. (1999). Performance measurements on sandglass-type parallelization of doacross loops. In: Sloot, P., Bubak, M., Hoekstra, A., Hertzberger, B. (eds) High-Performance Computing and Networking. HPCN-Europe 1999. Lecture Notes in Computer Science, vol 1593. Springer, Berlin, Heidelberg . https://doi.org/10.1007/BFb0100627
Download citation
DOI: https://doi.org/10.1007/BFb0100627
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-65821-4
Online ISBN: 978-3-540-48933-7
eBook Packages: Springer Book Archive