Po-Jen Chuang
Abstract
Fully recognizing various subcubes in a hypercube computer efficiently is nontrivial due to the specific structure of the hypercube. We propose a method with much less complexity than the multiple-GC strategy in generating the search space, while achieving complete subcube recognition. This method is referred to as a dynamic processor allocation scheme because the search space generated is dependent upon the dimension of the requested subcube dynamically, rather than being predetermined and fixed. The basic idea of this strategy lies in collapsing the binary tree representations of a hypercube successively so that the nodes which form a subcube but are distant would be brought close to each other for recognition. The strategy can be implemented efficiently by using shuffle operations on the leaf node addresses of binary tree representations. Extensive simulation runs are carried out to collect experimental performance measures of interest of different allocation strategies. It is shown from analytic and experimental results that this strategy compares favorably in many situations to any other known allocation scheme capable of achieving complete subcube recognition.
- 1 Y. Ssad and M. I-f. Sehults, "Topological Properties of Hypercubes," IEEE Trans. on Comp,uiers, pp. 867-872, July 1988. Google Scholar
Digital Library
- 2 C. L. S&b, The Capic Cube, CACM, vol. 28, No. 1, pp. 2923, Jan. 1985.Google Scholar
- 3 Intel Corporation, A New Direction in Scientific Cotnpufiag, Order #28QOSQQl, Intel Corporation, 1985.Google Scholar
- 4 W. D. Hillis, The Connection Machine, Cambridge, MA The MIT Press, 1985. Google ScholarDigital Library
- 5 Ametek System 14 User's Guide: C Edition, Ametek Computer Resesreh Division, Arcadia, California, 1988.Google Scholar
- 6 J. P. Hayes et al., A Microproeesscr-Based Hypercube Supercomputer, IEEE Micro, vol. 8, pp. 6-17, Oct. 1988. Google ScholarDigital Library
- 7 H. L. Gust&on, S. Hawkinson, and K. Scott, The Archi&- ture of a Homogepeous V&or Supercomputer, Proe. 1986 In17 Con/ Par&l Processing, Aug. 1986, pp. 649652.Google Scholar
- 8 J. C. Peterson CC al., The Mark LB Hypercube-Ensemble Concurrent Computer, Proc. 1985 Int'l Conf ParaLl Proceasing, Aug. 1985, pp. 71-73.Google Scholar
- 9 M.-S. Chen and K. G. Shin, Processor Allocation in an N- Cube Multiprocessor Using Gray Codes, IEEE Trans. on Compufers, O-36, 12, pp. 13051407, Dec. 1987. Google ScholarDigital Library
- 10 K. C. Knowlton, A Fast Storage Allocator, CACM, vol. 8, No. 10, pp. 623425, Oct. 1966. Google ScholarDigital Library
- 11 H. S. Stone, Paraliel Pmcessing with the Perfect Shuffle, IEEE Trans. on Compufcra, vol. C20, pp. 153-161, Feb. 1971.Google ScholarDigital Library
- 12 M.-S. Chen snd K. G. Shin, "Task Migration in Hypercubc Multipmcesors, Proc. fbth Annual Int't S'ymp. Computer Arehitetturc, May 1989, pp. 105LlL. Google ScholarDigital Library
- 13 A. Al-Dhelaan sad B. Bose, A New Strategy for Processor Allocation in au N-cube Multiprocessor, Proe. Phocniz Conf Uomp. and Oomm., Mar. 1989, pp. 114118.Google Scholar
- 14 J. Kim, C. R. Dss, and W. Lin, A Processor Allocation Scheme for Hypercube Computers, Proc. 1989 Int'l Cont. Parallel Processing, Aug. 1989, pp. II 231338.Google Scholar
Index Terms
- Dynamic processor allocation in hypercube computers
Recommendations
Dynamic processor allocation in hypercube computers
ISCA '90: Proceedings of the 17th annual international symposium on Computer ArchitectureFully recognizing various subcubes in a hypercube computer efficiently is nontrivial due to the specific structure of the hypercube. We propose a method with much less complexity than the multiple-GC strategy in generating the search space, while ...
A Fast Recognition-Complete Processor Allocation Strategy for Hypercube Computers
Fully recognizing various subcubes in a hypercube computer efficiently is addressed. A method with much less complexity than the multiple-GC strategy in generating the search space, while achieving complete subcube recognition, is proposed. This method ...
Processor Allocation in Hypercube Multiprocessors
The processor allocation problem requires recognizing and locating a free subcube that can accommodate a request for a subcube of a specified size for an incoming task. Methods reported in the literature fall into two strategies: Bottom-up or bit mapped ...
Comments