Abstract
A very brief introduction to interconnection networks is given. It is then shown that, in a restricted sense, an ideal N-by-N crossbar switch can be simulated with slowdown factor logdN by a fully extended omega network built from d-by-d switches. Next, lower bounds are given for the product of the average traffic per link and the number of network links. The bounds are valid for interconnection networks with an upperbound on the number of links coming out of each node. Finally, it is shown that in a probabalistic sense, the peak traffic per link can be made close to the bound by using a slightly extended omega network. Randomized routing plays a key role in reducing congestion.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Alt U, Hagerup T, Mehlhorn K, Preparata FP (1986) Deterministic simulation of idealized parallel computers on more realistic ones.
Chin C-Y, Hwang K (1984) Packet switching networks for multiprocessors and dataflow computers, IEEE Transactions on Computers 33:991–1003
Cruz (1987) Maximum delay in interconnection networks. Forthcoming Ph. D. thesis
Greenberg RL, Leiserson CE (1985) Randomized routing on fat trees. Proc 26th IEEE Annual Symp on Foundations of Computer Science 26:241–249
Hoeffding W (1963) Probability inequalities for sums of bounded random variables. Journal American Statistical Society 58:13–30
Hwang K (ed)(1984) Tutorial: supercomputers—design and applications. IEEE Computer Society Press 648pp
Kermani P, Kleinrock L (1979) Virtual cut-through: a new computer communication switching technique. Computer Networks 3:267–286
Kermani P, Kleinrock L (1980) A tradeoff study of switching systems in computer communication networks. IEEE Transactions on Computers 29:1052–1060
Lawrie DH (1975) Access and alignment of data in an array processor. IEEE Transactions on Computers 24:1145–1155
Mitra D, Cieslak R (1986) Randomized parallel communications on an extension of the omega network. Proc 1986 Int Conf on Parallel Processing
Pippenger N (1984) Parallel communication with limited buffers. Proc 25th IEEE Annual Symp on Foundations of Computer Science 25:127–136.
Pippenger N, Valiant LG (1976) Shifting graphs and their applications. Journal of the Association for Computing Machinery 23:423–432
Raghaven P, Thompson CD (1985) Provably good routing in graphs: regular arrays. Proc ACM Symposium on Theory of Computing 17:79–87
Shannon CE (1950) Memory requirements in a telephone exchange. Bell System Technical Journal 29:343–349
Upfal E (1984) Efficient schemes for parallel communication. Journal of the Association for Computing Machinery 31:507–517
Valiant LG (1982) A scheme for fast parallel communication. SIAM J. Computing 11:350–361
Wu C- l, Feng Ty (eds) (1984) Tutorial: interconnection networks for parallel and distributed processing. IEEE Computer Society Press 656pp
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1987 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Hajek, B., Cruz, R.L. (1987). Delay and Routing in Interconnection Networks. In: Odoni, A.R., Bianco, L., Szegö, G. (eds) Flow Control of Congested Networks. NATO ASI Series, vol 38. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-86726-2_14
Download citation
DOI: https://doi.org/10.1007/978-3-642-86726-2_14
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-86728-6
Online ISBN: 978-3-642-86726-2
eBook Packages: Springer Book Archive