A simple and efficient input selection function for networks-on-chip

N/ACitations
Citations of this article
3Readers
Mendeley users who have this article in their library.
Get full text

Abstract

Wormhole-switching and virtual channel flow control are two critical techniques in networks-on-chip (NoCs). In an NoC adopting these two techniques, a packet may hold several virtual channel (vc) resources spanning multiple routers. These vcs constitute a vc chain to the packet. Through observation, we find that the lengths of the vc chains play an important role in the performance of an NoC, and it helps to improve the performance of the network to cut short the vc chains. In this paper, we propose a novel input selection function (ISF) which allows packets spanning in the network in a more compact and consecutive manner, thereby lowering the delay while simultaneously boosting throughput. Owing to the simplicity of the novel ISF, we can implement it with a practical design, incurring a minimal hardware overhead with an additional requirement of storage less than 3.6%. We simulate and evaluate the proposed input selection approach in terms of average delay and throughput. Our experimental results indicate that the proposed ISF is effective in NoC design compared to other ISFs in previous literatures. Though we assume a two-dimensional mesh topology throughout this paper, the proposed ISF can be readily extended to other topologies. Furthermore, it can be coupled with any OSF and any routing algorithm. © 2012 Springer-Verlag.

Cite

CITATION STYLE

APA

Wang, X., Yu, Z., & Xu, H. (2012). A simple and efficient input selection function for networks-on-chip. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7129 LNCS, pp. 525–539). https://doi.org/10.1007/978-3-642-25959-3_39

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free