Floorplanning and Evaluation of A Hierarchical Cluster-Based NoC for Multicore System

The WK-recursive network well conforms to implementation of a large distributed system due to its many favorable properties such as regularity and recursive scalability. In this paper, we propose a hierarchical cluster-based NoC that blends WK-recursive network and Pyramid WK-recursive network, where each sub-network is optimized for different traffic types. To better exploit the communication locality of the asymmetric topology, we present a specific implementation with tile shapes and Manhattan-style wire. Using a extended cycle-accurate simulator Noxim, the experiment results show that our proposed design provides 18.6%–29.3% improvement in average network latency and 11.2%–17.8% improvement in average network throughput over the previous architecture evaluated in this study. It is demonstrated that it is feasible to implement the triangular topologies with rectangular manhattan architecture.

[1]  Henry Hoffmann,et al.  On-Chip Interconnection Architecture of the Tile Processor , 2007, IEEE Micro.

[2]  J. Gonzalez,et al.  A New Bus Assignment Algorithm for a Shared Bus Switch Fabric , 2000, VLSI Design.

[3]  Martin Hopkins,et al.  Synergistic Processing in Cell's Multicore Architecture , 2006, IEEE Micro.

[4]  Jason Howard A 48-core IA-32 processor with on-die message-passing and DVFS in 45nm CMOS , 2010, 2010 IEEE Asian Solid-State Circuits Conference.

[5]  Feng Shi,et al.  A Triplet-based Computer Architecture Supporting Parallel Object Computing , 2007, 2007 IEEE International Conf. on Application-specific Systems, Architectures and Processors (ASAP).

[6]  Onur Mutlu,et al.  Kilo-NOC: A heterogeneous network-on-chip architecture for scalability and service guarantees , 2011, 2011 38th Annual International Symposium on Computer Architecture (ISCA).

[7]  William J. Dally,et al.  Flattened Butterfly Topology for On-Chip Networks , 2007, IEEE Comput. Archit. Lett..

[8]  Onur Mutlu,et al.  Express Cube Topologies for on-Chip Interconnects , 2009, 2009 IEEE 15th International Symposium on High Performance Computer Architecture.

[9]  Jaehyuk Huh,et al.  Exploiting ILP, TLP, and DLP with the Polymorphous TRIPS Architecture , 2003, IEEE Micro.

[10]  Yang Zhang,et al.  Floorplanning exploration and performance evaluation of a new Network-on-Chip , 2011, 2011 Design, Automation & Test in Europe.

[11]  Philip Heidelberger,et al.  Blue Gene/L torus interconnection network , 2005, IBM J. Res. Dev..