A NoC Traffic Suite Based on Real Applications

As benchmark programs for microprocessor architectures, network-on-chip (NoC) traffic patterns are essential tools for NoC performance assessments and architecture explorations. The fidelity of NoC traffic patterns has profound influence on NoC studies. For the first time, this paper presents a realistic traffic benchmark suite, called MCSL, and the methodology used to generate it. The publicly released MCSL benchmark suite includes a set of realistic traffic patterns for 8 real applications and covers popular NoC architectures. It captures not only the communication behaviors in NoCs but also the temporal dependencies among them. MCSL benchmark suite can be easily incorporated into existing NoC simulators and significantly improve NoC simulation accuracy. We developed a systematic traffic generation methodology to create MCSL based on real applications. The methodology uses formal computational models to capture both communication and computation requirements of applications. It optimizes application mapping and scheduling to faithfully maximize overall system performance and utilization before extracting realistic traffic patterns through cycle-accurate simulations. Experiment results show that MCSL benchmark suite can be used to study NoC characteristics more accurately than traditional random traffic patterns.

[1]  W. Dally,et al.  Route packets, not wires: on-chip interconnection networks , 2001, Proceedings of the 38th Design Automation Conference (IEEE Cat. No.01CH37232).

[2]  William J. Dally,et al.  Route packets, not wires: on-chip inteconnection networks , 2001, DAC '01.

[3]  Hironori Kasahara,et al.  A standard task graph set for fair evaluation of multiprocessor scheduling algorithms , 2002 .

[4]  Soonhoi Ha,et al.  H.264 Decoder Algorithm Specification and Simulation in Simulink and PeaCE , 2004 .

[5]  Radu Marculescu,et al.  On-chip traffic modeling and synthesis for MPEG-2 video applications , 2004, IEEE Transactions on Very Large Scale Integration (VLSI) Systems.

[6]  Jörg Henkel,et al.  H. 264 HDTV Decoder Using Application-Specific Networks-On-Chip , 2005, 2005 IEEE International Conference on Multimedia and Expo.

[7]  Tobias Bjerregaard,et al.  A survey of research and practices of Network-on-chip , 2006, CSUR.

[8]  Li-Shiuan Peh,et al.  A Statistical Traffic Model for On-Chip Interconnection Networks , 2006, 14th IEEE International Symposium on Modeling, Analysis, and Simulation.

[9]  Jinwen Xi,et al.  A System-level Network-on-Chip Simulation Framework Integrated with Low-level Analytical Models , 2006, 2006 International Conference on Computer Design.

[10]  T.Z. Islam,et al.  Gpnocsim - A General Purpose Simulator for Network-On-Chip , 2007, 2007 International Conference on Information and Communication Technology.

[11]  Gerard J. M. Smit,et al.  Fast, Accurate and Detailed NoC Simulations , 2007, First International Symposium on Networks-on-Chip (NOCS'07).

[12]  Radu Marculescu,et al.  An Initiative towards Open Network-on-Chip Benchmarks , 2007 .

[13]  J. Bahn,et al.  A Generic Traffic Model for On-Chip Interconnection Networks , 2008 .

[14]  L. Anghel,et al.  A flexible network-on-chip simulator for early design space exploration , 2008, 2008 1st Microsystems and Nanoelectronics Research Conference.

[15]  Xiaowen Wu,et al.  Satisfiability Modulo Graph Theory for Task Mapping and Scheduling on Multiprocessor Systems , 2011, IEEE Transactions on Parallel and Distributed Systems.