On modelling of an adaptive routing algorithm for massively parallel processing systems

Abstract In this work, the message latency of an adaptive routing algorithm for a hypercube circuit switching interconnection network is analyzed. This routing algorithm is based on back-track-to-the-origin-and-retry approach. Intermediate links are chosen among the ones that form the shortest paths from a source to the destination, after excluding all the allocated links, the algorithm selects one of the remaining free links at random. The subsequent link in the path is established through the selected link and the procedure is repeated until the destination is reached, or when no free links can be found. If no free links are to be found at an intermediate node in the path, then a break is returned to the origin, the partial path is dissolved, and after a back-off period the path set up phase is repeated. The average message latency for such a protocol is here determined using an iterative approximated analytical model; peculiarity of the model is that it takes into account the feedback effect between the probability of conflict to obtain a link and the message latency itself. An accurate simulation analysis has been carried out to validate the proposed model. Finally the adaptive protocol is compared with the e-cube based protocol currently used by commercial products. The derived results show a better behaviour of the previous ones independent of the message length distribution.

[1]  Dharma P. Agrawal,et al.  Performance of multiprocessor interconnection networks , 1989, Computer.

[2]  Krishnan Padmanabhan,et al.  Performance of the Direct Binary n-Cube Network for Multiprocessors , 1989, IEEE Trans. Computers.

[3]  Mary K. Vernon,et al.  Performance Analysis of Mesh Interconnection Networks with Deterministic Routing , 1994, IEEE Trans. Parallel Distributed Syst..

[4]  D. Grunwald,et al.  The Performance of Multicomputer Interconnection Networks , 1987, Computer.

[5]  Charles L. Seitz,et al.  The cosmic cube , 1985, CACM.

[6]  E. Chow,et al.  A high-speed message-driven communication architecture , 1988, ICS '88.

[7]  Richard M. Fujimoto,et al.  Multicomputer Networks: Message-Based Parallel Processing , 1987 .

[8]  Duncan H. Lawrie,et al.  Performance analysis of redundant-path networks for multiprocessor systems , 1985, TOCS.

[9]  Abraham Silberschatz,et al.  Operating System Concepts , 1983 .

[10]  Chita R. Das,et al.  Hypercube Communication Delay with Wormhole Routing , 1994, IEEE Trans. Computers.

[11]  Stephen S. Lavenberg,et al.  Computer Performance Modeling Handbook , 1983, Int. CMG Conference.

[12]  Gregory V. Wilson,et al.  A glossary of parallel computing terminology , 1993, IEEE Parallel & Distributed Technology: Systems & Applications.

[13]  Daniel H. Linder,et al.  An Adaptive and Fault Tolerant Wormhole Routing Strategy for k-Ary n-Cubes , 1994, IEEE Trans. Computers.

[14]  Peter G. Harrison,et al.  The representation of multistage interconnection networks in queuing models of parallel systems , 1990, JACM.

[15]  Andrew A. Chien,et al.  Planar-adaptive routing: low-cost adaptive networks for multiprocessors , 1992, ISCA '92.

[16]  Lionel M. Ni,et al.  A survey of wormhole routing techniques in direct networks , 1993, Computer.

[17]  Philip S. Yu,et al.  Analysis of Concurrency-Coherency Control Protocols for Distributed Transaction Processing Systems with Regional Locality , 1992, IEEE Trans. Software Eng..

[18]  Ming-Syan Chen,et al.  Adaptive Fault-Tolerant Routing in Hypercube Multicomputers , 1990, IEEE Trans. Computers.

[19]  Leonard Kleinrock,et al.  Virtual Cut-Through: A New Computer Communication Switching Technique , 1979, Comput. Networks.

[20]  Wade H. Shaw,et al.  A performance study of the hypercube parallel processor architecture , 1991, Simul..

[21]  Kang G. Shin,et al.  Traffic Routing for Multicomputer Networks with Virtual Cut-Through Capability , 1992, IEEE Trans. Computers.

[22]  Rupert G. Miller.  The jackknife-a review , 1974 .

[23]  S. F. Nugent,et al.  The iPSC/2 direct-connect communications technology , 1988, C3P.

[24]  Israel Koren,et al.  Discrete and Continuous Models for the Performance of Reconfigurable Multistage Systems , 1991, IEEE Trans. Computers.

[25]  Bruno Ciciani,et al.  On modeling link conflict resolution strategies for circuit-switching hypercubes , 1991, [1991] Proceedings, Advanced Computer Technology, Reliable Systems and Applications.

[26]  William J. Dally,et al.  Performance Analysis of k-Ary n-Cube Interconnection Networks , 1987, IEEE Trans. Computers.

[27]  Lionel M. Ni,et al.  The Turn Model for Adaptive Routing , 1992, [1992] Proceedings the 19th Annual International Symposium on Computer Architecture.

[28]  Richard J. Enbody,et al.  Performance evaluation of mesh-connected wormhole-routed networks for interprocessor communication in multicomputers , 1990, Proceedings SUPERCOMPUTING '90.

[29]  Anant Agarwal,et al.  Limits on Interconnection Network Performance , 1991, IEEE Trans. Parallel Distributed Syst..

[30]  Chuan-lin Wu,et al.  Performance Analysis of Multistage Interconnection Network Configurations and Operations , 1992, IEEE Trans. Computers.

[31]  Philip S. Yu,et al.  Analysis of Replication in Distributed Database Systems , 1990, IEEE Trans. Knowl. Data Eng..

[32]  Anthony J. G. Hey.  Supercomputing with transputers—past, present and future , 1990, ICS '90.

[33]  Krishnan Padmanabhan.  Cube structures for multiprocessors , 1990, CACM.

[34]  N.J. Dimopoulos,et al.  Performance evaluation of the backtrack-to-the-origin-and retry routing for hypercycle-based interconnection networks , 1990, Proceedings.,10th International Conference on Distributed Computing Systems.