
This Article  
 
Share  
Bibliographic References  
Add to:  
Digg Furl Spurl Blink Simpy Del.icio.us Y!MyWeb  
Search  
 
ASCII Text  x  
G. Karypis, V. Kumar, "Unstructured Tree Search on SIMD Parallel Computers," IEEE Transactions on Parallel and Distributed Systems, vol. 5, no. 10, pp. 10571072, October, 1994.  
BibTex  x  
@article{ 10.1109/71.313122, author = {G. Karypis and V. Kumar}, title = {Unstructured Tree Search on SIMD Parallel Computers}, journal ={IEEE Transactions on Parallel and Distributed Systems}, volume = {5}, number = {10}, issn = {10459219}, year = {1994}, pages = {10571072}, doi = {http://doi.ieeecomputersociety.org/10.1109/71.313122}, publisher = {IEEE Computer Society}, address = {Los Alamitos, CA, USA}, }  
RefWorks Procite/RefMan/Endnote  x  
TY  JOUR JO  IEEE Transactions on Parallel and Distributed Systems TI  Unstructured Tree Search on SIMD Parallel Computers IS  10 SN  10459219 SP1057 EP1072 EPD  10571072 A1  G. Karypis, A1  V. Kumar, PY  1994 KW  Index Termsresource allocation; parallel machines; trees (mathematics); search problems; parallelarchitectures; performance evaluation; unstructured tree search; SIMD parallelcomputers; load balancing; unstructured tree computations; largescale SIMD machines;scalability; tree search; SIMD machine; triggering mechanism; search spaceredistribution; loadbalancing methods; loadbalancing schemes; MIMD architectures;15puzzle problem; CM2 SIMD parallel computer VL  5 JA  IEEE Transactions on Parallel and Distributed Systems ER   
We present new methods for load balancing of unstructured tree computations onlargescale SIMD machines, and analyze the scalability of these and other existingschemes. An efficient formulation of tree search on an SIMD machine consists of twomajor components: a triggering mechanism, which determines when the search spaceredistribution must occur to balance the search space over processors, and a scheme toredistribute the search space. We have devised a new redistribution mechanism and anew triggering mechanism. Either of these can be used in conjunction with triggering andredistribution mechanisms developed by other researchers. We analyze the scalability ofthese mechanisms and verify the results experimentally. The analysis and experimentsshow that our new loadbalancing methods are highly scalable on SIMD architectures.Their scalability is shown to he no worse than that of the best loadbalancing schemeson MIMD architectures. We verify our theoretical results by implementing the 15puzzleproblem on a CM2 SIMD parallel computer.
[1] S. Arvindam, V. Kumar, and V.N. Rao, "Efficient parallel algorithms for search problems: Applications in VLSI CAD," inProc. Frontiers 90 Conf. Massively Parallel Computation, Oct. 1990.
[2] S. Arvindam, V. Kumar, V.N. Rao, and V. Singh, "Automatic test Pattern generation on multiprocessors,"Parallel Computing, vol. 17, pp. 13231342, Dec. 1991.
[3] G. E. Blelloch, "Scans as primitive parallel operations,"IEEE Trans. Comput., vol. 11, pp. 15261538, 1989.
[4] M. Evett, J. Hendler, A. Mahanti, and D. Nau, "PRA*: A memorylimited heuristic search procedure for the connection machine," inProc. 3rd Symp. Frontiers of Massively Parallel Computation, 1990, pp. 145149.
[5] C. Ferguson and R. Korf, "Distributed tree search and its application to alphabeta pruning," inProc. 1988 Natl. Conf. Artificial Intell., Aug. 1988.
[6] R. Finkel and U. Manber, "DIBA distributed implementation of backtracking,"ACM Trans. Programming Languages Syst., vol. 9, no. 2, pp. 235256, Apr. 1987.
[7] R. Frye and J. Myczkowski, "Exhaustive search of unstructured trees on the connection machine," Tech. Rep., Thinking Machines Corp., Cambridge, MA. 1990.
[8] M. Furuichi, K. Taki, and N. Ichiyoshi, "A multilevel load balancing scheme for ORparallel exhaustive search programs on the multiPSI," inProc. 2nd ACM SIGPLAN Symp. Principles and Practice of Parallel Programming, 1990, pp. 5059.
[9] A. Grama, V. Kumar, and V. N. Rao, "Experimental evaluation of load balancing techniques for the hypercube," inProc. Parallel Computing 91 conf., 1991.
[10] A. Gupta and V. Kumar, "On the Scalability of FFT on Parallel Computers,"Proc. Frontiers 90 Conf. Massively Parallel Computation, 1990.
[11] J. L. Gustafson, "Reevaluating Amdahl's Law,"Commun. ACM, vol. 31, no. 5, pp. 532533, 1988.
[12] J. L. Gustafson, G. R. Montry, and R. E. Benner, "Development of parallel methods for a 1024processor hypercube,"SIAM J. Sci. Statistical Computing, vol. 9, pp. 609638, 1988.
[13] D. Hillis,The Connection Machine. Cambridge, MA: M.I.T. Press, 1985.
[14] E. Horowitz and S. Sahni,Fundamentals of Computer Algorithms. Rockville, MD: Computer Sci. Press, 1978.
[15] L. Kanal and V. Kumar,Search in Artificial Intelligence. New York: SpringerVerlag. 1988.
[16] G. Karypis and V. Kumar, "Unstructured tree search on SIMD parallel computers: A summary of results,"Supercomputing, 1992.
[17] R. E. Korf, "Depthfirst iterative deepening: An optimal admissible tree search,"Artificial Intell., vol. 25, pp. 97109, 1985.
[18] V. Kumar, "DepthFirst Search," in S. C. Shapiro, Ed.,Encyclopaedia of Artificial Intelligence, vol. 2. New York: Wiley, 1987, pp. 10041005.
[19] V. Kumar et al.,Introduction to Parallel Computing: Design and Analysis of Parallel Algorithms, Benjamin Cummings, Redwood City, Calif., 1994.
[20] V. Kumar, A. Grama, and V. N. Rao, "Scalable load balancing techniques for parallel computers,"J. Parallel Distrib. Computing, vol. 22, pp. 6079, 1994.
[21] V. Kumar and A. Gupta, "Analyzing scalability of parallel algorithms and architectures,"J. parallel Distrib. Computing, 1994.
[22] V. Kumar, D. Nau, and L. Kanal,General branchandbound formulation for AND/OR graph and game tree search, in L. Kanal and V. Kumar, Eds.,Search in Artificial Intelligence. New York: SpringerVerlag, 1988.
[23] V. Kumar and V. N. Rao, "Parallel depthfirst search, Part II: Analysis,"Int. J. Parallel Program., vol. 16, no. 6, pp. 501519, 1987.
[24] V. Kumar and V. Singh, "Scalability of Parallel Algorithms for the AllPairs ShortestPath Problem,"J. Parallel and Distributed Computing, Vol. 13, No. 2, Oct. 1991, pp. 124138.
[25] R. M. Karp, "Challenges in combinatorial computing," to appear.
[26] A. Mahanti and C. Daniels, "SIMD parallel heuristic search,"Artificial Intell., 1992.
[27] B. Monien and O. Vornberger, "Parallel processing of combinatorial search trees," inProc. Int. Workshop on Parallel Algorithms and Architectures, May 1987.
[28] N. Nilsson,Principles of Artificial Intelligence. Palo Alto, CA: Tioga, 1980.
[29] C. H. Papadimitriou and K. Steiglitz,Combinatorial Optimization: Algorithms and Complexity. Englewood Cliffs, NJ: PrenticeHall, 1982.
[30] S. Patil and P. Banerjee, "A parallel branch and bound approach to test generation,"IEEE Trans. Comput.Aided Design of Circuits and Syst., vol. 9, pp. 313322, Mar. 1990.
[31] J. Pearl,Heuristics: Intelligent Search Strategies for Computer Problem Solving. Reading, Mass: AddisonWesley, 1984.
[32] C. Powley, R.E. Korf, and C. Ferguson, "IDA*on the connection machine,"Artificial Intell., 1992.
[33] C. Powley and R. E. Korf, "SIMD and MIMD parallel search," inProc. AAAI Spring Symp., 1989, pp. 4953.
[34] A. G. Ranade, "Optimal speedup for backtrack search on a Butterfly network," inProc. 3rd ACM Symp. Parallel Algorithms and Architectures, 1991.
[35] Ranka, S., and S. Sahni,Hypercube Algorithms for Image Processing and Pattern Recognition, SpringerVerlag, Berlin, 1990.
[36] V. N. Rao and V. Kumar, "Parallel depthfirst search, Part I: Implementation,"Int. J. Parallel Programming, vol. 16, no. 6, pp. 479499, 1987.
[37] V.N. Rao and V. Kumar, "On the efficiency of parallel backtracking,"IEEE Trans. Parallel Distrib. Syst., vol. 4, pp. 427437, Apr. 1993.
[38] J. Myczkowski and R. Frye, "Exhaustive search of unstructured trees on the connection machine," Tech. Rep., Thinking Machines Corp., 1990.
[39] J. Myczkowski and R. Frye, "Load balancing algorithms on the connection machine and their use in MonteCarlo methods," inProc. Unstructured Sci. Computation on Multiprocessors Conf., 1992.
[40] V. Saletore and L. V. Kale, "Consistent linear speedup to a first solution in parallel statespace search," inProc. 1990 Natl. Conf. Artificial Intell., 1990, pp. 227233.
[41] W. Shu and L. V. Kale, "A dynamic scheduling strategy for the Charekernel system," inProc. ACM Supercomput. Conf., 1989, pp. 389398.
[42] V. Singh, G. Agha, V. Kumar, and C. Tomlinson, "Scalability of parallel sorting on mesh multicomputers," inProc. 5th Int. Parall. Processing Symp., 1991.
[43] B. W. Wah, G. J. Li, and C. F. Yu, "Multiprocessing of combinatorial search problems,"IEEE Comput., June 1985.
[44] B. W. Wah and E. Ma, "MANIP: A multicomputer architecture for solving combinatorial extremumsearch problems,"IEEE Trans. Comput., vol. C33, no. 5, pp. 377390, May 1984.
[45] J. Woo and S. Sahni, "Computing biconnected components on a hypercube,"J. Supercomputing, June 1991.
[46] J. Woo and S. Sahni, "Computing biconnected components on a hypercube,"J. Supercomputing, June 1991.