A power-efficient network on-chip topology

NoCs have become a critical component in many-core architectures. Usually, the preferred topology is the 2D-Mesh as it enables a tile-based layout significantly reducing the design effort. However, new emerging challenges such as power consumption need to be addressed. Looking at the NoC, routers and links not being used must be switched off, thus achieving large power savings. Topology and routing algorithm must be carefully designed as they may lack enough flexibility to switch off components for long periods of time. We present the NR-Mesh (Nearest neighboR Mesh) topology. It gives an end node the choice to inject a message through different neighboring routers, thereby reducing hop count and saving latency. At the receiver side, a message may be delivered to the end node through different routers, thus reducing hop count further and increasing flexibility. When allowing links and routers to switch off and combined with adaptive routing, the power management technique is able to achieve significant power savings (up to 36% savings in static power consumed at routers). When compared with the 2D-Mesh, NR-Mesh reduces execution time by 23% and power consumption at routers by 47%.

[1]  Min Xu,et al.  Evaluating Non-deterministic Multi-threaded Commercial Workloads , 2001 .

[2]  Andrew B. Kahng,et al.  ORION 2.0: A fast and accurate NoC power and area model for early-stage design space exploration , 2009, 2009 Design, Automation & Test in Europe Conference & Exhibition.

[3]  William J. Dally,et al.  Flattened Butterfly Topology for On-Chip Networks , 2007, 40th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO 2007).

[4]  William J. Dally,et al.  Express Cubes: Improving the Performance of k-Ary n-Cube Interconnection Networks , 1989, IEEE Trans. Computers.

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

[6]  Saurabh Dighe,et al.  An 80-Tile 1.28TFLOPS Network-on-Chip in 65nm CMOS , 2007, 2007 IEEE International Solid-State Circuits Conference. Digest of Technical Papers.

[7]  Mahmood Fathy,et al.  A Fault Tolerant NoC Architecture for Reliability Improvement and Latency Reduction , 2009, 2009 12th Euromicro Conference on Digital System Design, Architectures, Methods and Tools.

[8]  Fredrik Larsson,et al.  Simics: A Full System Simulation Platform , 2002, Computer.

[9]  Anoop Gupta,et al.  The SPLASH-2 programs: characterization and methodological considerations , 1995, ISCA.

[10]  M. Suzuoki,et al.  Overview of the architecture, circuit design, and physical implementation of a first-generation cell processor , 2006, IEEE Journal of Solid-State Circuits.

[11]  Wen-Hsiang Hu,et al.  DMesh : a Diagonally-Linked Mesh Network-on-Chip Architecture , 2008 .

[12]  William J. Dally,et al.  Design tradeoffs for tiled CMP on-chip networks , 2006, ICS '06.

[13]  Vivek Sarkar,et al.  Baring It All to Software: Raw Machines , 1997, Computer.

[14]  Milo M. K. Martin,et al.  Multifacet's general execution-driven multiprocessor simulator (GEMS) toolset , 2005, CARN.

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

[16]  C. Sanges,et al.  A recursively scalable network VLSI implementation , 1988, Future Gener. Comput. Syst..

[17]  Pradip Bose,et al.  Microarchitectural techniques for power gating of execution units , 2004, Proceedings of the 2004 International Symposium on Low Power Electronics and Design (IEEE Cat. No.04TH8758).