ABSTRACT
The Midimew network is an excellent contender for implementing the communication subsystem of a high performance computer. This network is an optimal 2D topology in the sense there are no other symmetric direct networks of degree 4 with a lower average distance or diameter. In fact, it reduces the diameter of the well known torus network by approximately □2. Although the topology was proposed and analyzed a decade ago, the lack of simple deadlock avoidance mechanisms prevented its utilization up to date. This study solved this drawback by applying the Bubble switching mechanism, a low cost deadlock-avoidance strategy developed by the authors. Moreover, by using routing tables we can configure our Virtual Cut-Through adaptive router to implement either a torus or a Midimew network. Thus, we can exploit the topological advantages of Midimew networks by simply changing the disposition of the wrap-around connections of its torus counterpart, without increasing the network implementation cost. To prove this assertion, we have carried out a thorough evaluation, from the hardware cost of the router to the parallel system performance under real loads.
- 1.A. Agarwal, "Limits on Interconnection Network Performance", IEEE Trans. on Comp., vol. 2, no4, pp:398- 412, October 1991. Google ScholarDigital Library
- 2.R. Beivide, E. Herrada, J. L. Balcazar, A. Armabarrena, "Optimal Distance Networks of Low Degree for Parallel Computers", IEEE Trans. on Comp., vol. 40, no. 10, pp. 1109-1123 November, 1991. Google ScholarDigital Library
- 3.A. Chien, "A cost and Speed Model for k-ary n-cube wormhole router", In Proc. of Hot Interconnects, August 1993.Google Scholar
- 4.W. J. Dally and C. L. Seitz, "Deadlock-Free Message Routing in Multiprocessor Interconnection Networks", IEEE Trans. on Comp., vol. C-36, 5, pp. 547-553, 1987. Google ScholarDigital Library
- 5.W.J. Dally, "Performance Analysis of k-ary n-cube Interconnection Networks", IEEE Trans. On Comp., Vol 39,No. 6 pp. 775-785, June 1990. Google ScholarDigital Library
- 6.M.Galles, "Scalable Pipelined Interconnect for Distributed Endpoint Routing: The SGI Spider Chip", Proc. of Hot Interconnects IV, August 1996.Google Scholar
- 7.J.R. Jump. "NETSIM Reference Manual". Rice University Electrical and Computer Engineering Department, March 1993.Google Scholar
- 8.C.M. Lau and G. Chen, "Optimal Layouts of Midimew Networks", IEEE Trans. on Parallel and Distributed Systems vol. 7, no. 9, pp. 954-961, September 1996. Google ScholarDigital Library
- 9.J. Laudon and D. Lenoski, "The SGI Origin: A cc-NUMA Highly Scalable Server", Proceedings of the 24th Annual International Symposium on Computer Architecture (ISCA-97), June 1997. Google ScholarDigital Library
- 10.J Laudon, K. Gharachorloo, W. Weber, A. Gupta, J. Hennessy, M. Horowitz, M.S. Lam "The Stanford DASH Multiprocessor" IEEE Computer no. 25 vol 3, pp- 63-79, March 1992. Google ScholarDigital Library
- 11.A. Nowatzyk, G. Aybay, M. Browne, E. Kelly, M. Parkin, B. Radke, and S. Vishin "The S3.mp Scalable Shared Memory Multiprocessor", Int Conf on Parallel Processing, August 1995.Google Scholar
- 12.V. S. Pai, P. Ranganathan, S. Adve "Rsim: An execution- Driven Simulator for ILP-Based Shared-Memory Multiprocessors and Uniprocessors", IEEE TCCA Newsletter, pp. 1-10 October. 1997. Google ScholarDigital Library
- 13.V. S. Pai, "RSIM Reference Manual. Version 1.0". Department of Electrical and Computer Engineering, Rice University. Technical Report 9705. July 1997.Google Scholar
- 14.J.M. Prellezo, V. Puente, J.A. Gregorio, R. Beivide, "SICOSYS: an into(connection network simulator for parallel computers," available at http:llwww.atc.unican.esl REPORTS/TR-ATC2-UC98.1xlf, June 1998.Google Scholar
- 15.V. Puente, J.A. Greg0rio, J. M. Prellezo, R.Beivide, J. Duato, and C. Izu, '!Adaptive Bubble Router: a Design to Balance Latency and Throughput in Networks for Parallel Computers", Proc. of International Conference on Parallel Computing, pp. 58-67, September. 1999. Google ScholarDigital Library
- 16.S.Scott and G. Thorson, "The Cray T3E Network: Adaptive Routing in a High Performance 3D Toms", Hot Interconnects IV, August 1996.Google Scholar
- 17.C.L.Seitz, "Concurrent VLSI architectures", IEEE Trans. on Comp., C-33, pp. 1247-1265, December 1984.Google ScholarDigital Library
- 18.S. C. Woo et al., "The SPLASH-2 Programs: Characterization andMethodological Considerations", In Proceedings of the~22nd International Symposium on Computer Architecture, pp. 24.36. June 1995. Google ScholarDigital Library
Index Terms
- Improving parallel system performance by changing the arrangement of the network links
Recommendations
On Improving the Performance of Hybrid Wired-Wireless Network-on-Chip Architectures
NoCArc '16: Proceedings of the 9th International Workshop on Network on Chip ArchitecturesRecently, hybrid wired-wireless Network-on-Chip (WiNoC) have been proposed to meet the performance and scalability demands of modern System-on-Chip (SoC) design. However, due to the presence of wirelines with multi-hop nodes in the hybrid architecture, ...
Rearranging links: a cost-effective approach to improve the reliability of multistage interconnection networks
The use of multiprocessor systems is the main method for providing a high computational power. Multistage interconnection networks MINs are widely used to connect processors and memory modules in multiprocessor systems. Therefore, the design of an ...
Comments