Temperature Management in 3D Network-on-Chips Using Simulated Annealing-Based Task Migration
Subject Areas : electrical and computer engineeringM. Mohebbi Moghaddam 1 * , S. H. Mir Mohammadi 2 , S. H. Mir Mohammadi 3
1 -
2 -
3 -
Keywords: 3D Network-on-chip temperature management task migration simulated annealing algorithm,
Abstract :
Combination of 3D stacking and network-on-chip (NoC), known as 3D NoC, has several advantages such as reduced propagation delay, chip area and interconnect, and power consumption, and bandwidth increase. Despite these advantages, 3D stacking causes the increased power density per chip area and subsequently increases the chip temperature. Temperature increase causes performance degradation and reliability reduction. Therefore, design of temperature management algorithms is essential for these systems. In this paper, we propose a task migration scheme for thermal management of 3D NoCs. The process of migration destinations for hot spots is an NP-complete problem which can be solved by using heuristic algorithms. To this end, we utilize a simulated annealing method in our algorithm. We consider migration overhead in addition to the temperature of the processing elements in migration destination selection process. Simulation results indicate up to 28 percentage peak temperature reduction, on average, for the benchmark that has the largest number of tasks. The proposed scheme has low migration overhead.
[1] C. Seiculescu, S. Murali, L. Benini, and G Micheli, "3D network on chipztopology synthesis: designing custom topologies for chip stacks," in 3D Integration for NoC-based SoC Architectures, Integrated Circuits and Systems, Springer New York, pp. 193-223, 2011.
[2] D. Jayasimha, B. Zafar, and Y Hoskote, On-Chip Interconnection Networks: Why They Are Different and How to Compare Them, Technical Report, Intel Corp, 2006.
[3] C. H. Chao, K. Y. Jheng, H. Y. Wang, J. C. Wu, and A. Y. Wu, "Traffic- and thermal-aware run-time thermal management scheme for 3D NoC systems," in Proc. 4th ACM/IEEE Int. Symp. on Networks-on-Chip, pp. 223-230, May 2010.
[4] Y. R. Huang, J. H. Pan, and Y. C. Lu, "Thermal-aware router-sharing architecture for 3D network-on-chip designs," in Proc. IEEE Asia Pacific Conf. on Circuits and Systems, pp. 1087-1090, 6-9 Dec. 2010.
[5] H. Wang, Y. Fu, T. Liu, and J. Wang, "Thermal management via task scheduling for 3D NoC based multi-processor," in Proc. Int. SoC Design Conf., ISOCC'10, pp. 440-444, 22-23 Nov. 2010.
[6] Y. Cui, W. Zhang, and H. Yu, "Distributed thermal-aware task scheduling for 3D network-on-chip," in Proc. 30th IEEE Int. Conf. on Computer Design, ICCD'12, pp. 494-495, 30 Sep.-3 Oct. 2012.
[7] C. Addo-Quaye, "Thermal-aware mapping and placement for 3-D NoC designs," in Proc. of the IEEE Int. Conf. on SOC, pp. 25-28, Sep. 2005.
[8] P. Hamedani, S. Hessabi, H. Sarbazi-Azad, and N. Enright Jerger, "Exploration of temperature constraints for thermal aware mapping of 3D networks on chip," International Journal of Adaptive, Resilient and Autonomic Systems, vol. 4, no. 3, pp. 42-60, Jul. 2012.
[9] J. Donald and M. Martonosi, "Techniques for multicore thermal management: classification and new exploration," in Proc. 33rd Int. Symp. on Computer Architecture, ISCA'06, pp. 78-88, Jun. 2006.
[10] F. Galea and R. Sirdey, "A parallel simulated annealing approach for the mapping of large process networks," in Proc. of 26th IEEE Int. Parallel and Distributed Processing Symp. Workshops & PhD Forum, IPDPSW'12, pp. 1787-1792, Shanghai, China, 21-25 May 2012.
[11] Z. Lu, L. Xia, and A. Jantsch, "Cluster-based simulated annealing for mapping cores onto 2D mesh networks on chip," in Proc. 11th IEEE Workshop on Design and Diagnostics of Electronic Circuits and Systems, DDECS'08, 6 pp., Apr. 2008.
[12] N. Hassanpour, S. Hessabi, and P. K. Hamedani, "Temperature control in three-network on chips using task migration," IET Computers & Digital Techniques, vol. 7, no. 6, pp. 274-281, Nov. 2013.
[13] H. Orsila, E. Salminen, and T. D. Hamalainen, "Best practices for simulated annealing in multiprocessor task distribution problems," in Simulated Annealing, Ch. 16, pp. 321-342, 2008.
[14] https://nocs.stanford.edu/cgi-bin/trac.cgi/wiki/Resources/BookSim, Accessed Mar. 2014.
[15] A. Kahng, B. Li, L. Peh, and K. Samadi, "ORION 2.0: a fast and accurate NoC power and area model for early-stage design space exploration," in Proc. Design, Automation & Test in Europe Conf. & Exhibition, DATE'09, pp. 423-428, Apr. 2009.
[16] A. Sharifi and H. Sarbazi-Azad, "Power consumption and performance analysis of 3D NoCs," in Proc. Asia-Pacific Computer Systems Architecture Conf., pp. 209-219, Aug. 2007.
[17] S. Priyadarshi, et al., "Hetero 3D integration: a scheme for optimizing efficiency/cost of chip multiprocessors," in Proc. 14th IEEE Int. Sym. on Quality Electronic Design, ISQED'13, 7 pp., 4-6 Mar. 2013.