A Novel 3D Mesh-Based NoC Architecture for Performance Improvement

Majlesi Journal of Electrical Engineering

مقاله 7، دوره 16، شماره 2، شهریور 2022، صفحه 85-101   

نوع مقاله: Review Article

شناسه دیجیتال (DOI): 10.30486/mjee.2022.696497

نویسندگان

Navid Habibi¹؛ MohammadReza Salehnamadi^{* 2}؛ Ahmad Khademzadeh³

¹Department of Computer Engineering, South Tehran Branch, Islamic Azad University, Tehran, Iran

²) Department of Computer Engineering, South Tehran Branch, Islamic Azad University, Tehran, Iran

³Telecommunication Research Center, Tehran, Iran

چکیده

Applying semiconductor technology, network-on-chips (NoCs) are designed on silicon chips to expand on-chip communications. Three-dimensional (3D) mesh-based architecture is also known as a basic NoC architecture characterized by better energy consumption and latency compared with two-dimensional (2D) ones.  Recently developed architectures are based on regular mesh. However, there are serious drawbacks in NoC architectures including high power consumption, energy consumption, and latency. Therefore, making an improvement in topology diameter would overcome these shortcomings. Accordingly, a new 3D mesh-based NoC architecture is proposed in the present study utilizing the star node, consisting of a new 3D topology with small diameter and new deadlock-free routing. The diameter of this architecture is then compared with its counterparts. Afterwards, the scalable universal matrix multiplication algorithm (SUMMA) is implemented in the proposed architecture. The results indicate a smaller network diameter, lower energy consumption (32%), less network latency (8.6%), as well as enhancement in throughput average (13.6%). The proposed matrix multiplication algorithm also implies improvement in the cost of the proposed architecture in comparison with its counterparts.

کلیدواژه‌ها

Communication؛ networks Architecture؛ Topology؛ Network-on-Chip؛ System-on-chip؛ Routing Protocols؛ Performance Evaluation؛ Multiplication Algorithm؛ Latency؛ Energy improvement؛ diameter

مراجع

[1] Feero, B. Stanley, & P. P. Pande, “Networks-on-chip in a three-dimensional environment: A performance evaluation,” IEEE Transactions on Computers, 2009.

[2] ABBAS, Assad, et al., “A survey on energy-efficient methodologies and architectures of network-on-chip,” Computers & Electrical Engineering, 2014.

[3] MANNA, Kanchan, et al., “Thermal-aware application mapping strategy for network-on-chip based system design,” IEEE Transactions on Computers, 2018.

[4] S. Azampanah, A. Khademzadeh, N. Bagherzadeh, M. Janidarmian, & R. Shojaee, “Contention-aware selection strategy for application-specific network-on-chip,” IET Computers & Digital Techniques, 2013.

[5] S. Forghani, N. Habibi, & M. Firoozbakht, “Network security metric based on attack duration,” In 2015 2nd IEEE International Conference on Knowledge-Based Engineering and Innovation (KBEI), 2015.

[6] A. More, V. Pano, & B. Taskin, “Vertical Arbitration-Free 3-D NoCs,” IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2018.

[7] R. K. Saini, & M. Ahmed, “2D hexagonal mesh Vs 3D mesh network on chip: A performance evaluation,” International Journal of Computing and Digital Systems, 2015.

[8] M. D. Schatz, R. A. Van de Geijn, & J. Poulson, “Parallel matrix multiplication: A systematic journey,” SIAM Journal on Scientific Computing, 2016.

[9] R. Dash, A. Majumdar, V. Pangracious, A. K. Turuk, & J. L. Risco-Martin, “ATAR: An Adaptive Thermal-Aware Routing Algorithm for 3-D Network-on-Chip Systems,” IEEE Transactions on Components, Packaging and Manufacturing Technology, 2018.

[10] R. Sabbaghi-Nadooshan, M. Modarressi, & H. Sarbazi-Azad, “A Novel De Bruijn Based Mesh Topology for Networks-on-Chip,” In VLSI. IntechOpen, BoD–Books on Demand, 2010.

[11] Y. Chen, J. Hu, X. Ling, & T. Huang, “A novel 3D NoC architecture based on De Bruijn graph,” Computers & Electrical Engineering, 2012.

[12] W. Gao, & P. Zhou, “Customized high performance and energy efficient communication networks for AI chips,” IEEE Access, 2019.

[13] R. Poovendran, & S. Sumathi, “An area‐efficient low‐power SCM topology for high performance network‐on Chip (NoC) architecture using an optimized routing design,” Concurrency and computation: practice and experience, 2019.

[14] N. Viswanathan, K. Paramasivam, & K. omasundaram, “Exploring Hierarchical, Cluster based 3D Topologies for 3D NoC,” Procedia Engineering, 2012.

[15] A. Touzene, “On all-to-all broadcast in dense Gaussian network on-chip,” IEEE Transactions on Parallel and Distributed Systems, 2015.

[16] I. Pires, M. Alves, & L. Albini, “Expansible Network-on-Chip Architecture,” Advances in Electrical and Computer Engineering, 2018.

[17] P. Yang, & Q. Wang, “Heterogeneous honeycomb-like NoC topology and routing based on communication division,” International Journal of Future Generation Communication and Networking, 2015.

[18] P. Lotfi-Kamran, A. M. Rahmani, M. Daneshtalab, A. Afzali-Kusha, & Z. Navabi, “EDXY–A low cost congestion-aware routing algorithm for network-on-chips,” Journal of Systems Architecture, 2010.

[19] N. Viswanathan, K. Paramasivam, & K. Somasundaram, “Performance and Cost Metrics Analysis of a 3D NoC Topology using Network Calculus,” Applied Mathematical Sciences, 2013.

[20] A. Ahmadinia, & A. Shahrabi, “A highly adaptive and efficient router architecture for network-on-chip,” The Computer Journal, 2011.

[21] D. Demirbas, I. Akturk, O. Ozturk, & U. Güdükbay, “Application-specific heterogeneous network-on-chip design,” The Computer Journal, 2014.

[22] S. E. Bae, T. W. Shinn, & T. Takaoka, “A faster parallel algorithm for matrix multiplication on a mesh array,” Procedia Computer Science, 2014.

[23] M. Hosseinabady, M. R. Kakoee, J. Mathew, & D. K. Pradhan, “Low latency and energy efficient scalable architecture for massive NoCs using generalized de Bruijn graph,” IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2011.

[24] O. Collins, F. Pollara, S. Dolinar, & J. Statman, “Wiring Viterbi decoders (splitting de Bruijn graphs),” The Telecommunications and Data Acquisition Progress Report, 1989.

[25] A. Louri, H. & Sung, “An efficient 3D optical implementation of binary de Bruijn networks with applications to massively parallel computing,” In IEEE Proceedings of Second International Workshop on Massively Parallel Processing Using Optical Interconnections, 1995.

[26] M. R. Samatham, & D. K. Pradhan, “The de Bruijn multiprocessor network: a versatile parallel processing and sorting network for VLSI,” IEEE Transactions on Computers, 1989.

[27] P. Faizian, M. A. Mollah, X. Yuan, Z. Alzaid, S. Pakin, & M. Lang, “Random Regular Graph and Generalized De Bruijn Graph with k-Shortest Path Routing,” IEEE Transactions on Parallel and Distributed Systems, 2018.

[28] D. Hoxha, “Sparse Matrices and Summa Matrix Multiplication Algorithm in STAPL Matrix Framework,” Doctoral dissertation, 2016.

[29] M. Schatz, J. Poulson, & R. van de Geijn, “Parallel Matrix Multiplication: 2D and 3D”, FLAME Working Note No. 62, The University of Texas at Austin, 2012.

[30] Y. S. Yang, H. Deshpande, G. Choi, & P. V. Gratz, “SDPR: Improving Latency and Bandwidth in On-Chip Interconnect Through Simultaneous Dual-Path Routing,” IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2018.

[31] T. Maqsood, S. Ali, S. U. Malik, & S. Madani, “A.Dynamic task mapping for network-on-chip based systems,” Journal of Systems Architecture, 2015.

[32] N. Jafarzadeh, M. Palesi, A. Khademzadeh, & A. Afzali-Kusha, “Data encoding techniques for reducing energy consumption in network-on-chip,” IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2014.

[33] H. Matsutani, M. Koibuchi, & H. Amano, “Tightly-coupled multi-layer topologies for 3-D NOCs,” In International Conference on Parallel Processing (ICPP), IEEE, 2007.

[34] K. Hwang, & N. Jotwani, “Advanced Computer Architecture,” 3e. McGraw-Hill Education, 2016.