一種基于自適應(yīng)容錯鏈路的片上網(wǎng)絡(luò)設(shè)計與研究

徐冬雨; 歐陽一鳴; 黃正峰; 李建華; 梁華國

doi:10.11999/JEIT240162

一種基于自適應(yīng)容錯鏈路的片上網(wǎng)絡(luò)設(shè)計與研究

doi: 10.11999/JEIT240162

1.
安徽工程大學(xué)計算機與信息學(xué)院蕪湖 241000
2.
合肥工業(yè)大學(xué)計算機與信息學(xué)院合肥 230009
3.
合肥工業(yè)大學(xué)微電子學(xué)院合肥 230009

基金項目: 國家自然科學(xué)基金( 62374049, 62174048, 62274052)

詳細信息

作者簡介:
徐冬雨：男，講師，研究方向為集成芯片與芯粒技術(shù)、可重構(gòu)片上互連網(wǎng)絡(luò)架構(gòu)

歐陽一鳴：男，教授，研究方向為片上系統(tǒng)與片上網(wǎng)絡(luò)

黃正峰：男，教授，研究方向為集成電路容錯設(shè)計

李建華：男，副教授，研究方向為計算機體系結(jié)構(gòu)、存儲系統(tǒng)、片上網(wǎng)絡(luò)

梁華國：男，教授，研究方向為數(shù)字系統(tǒng)設(shè)計自動化

通訊作者:
歐陽一鳴　oyymhfut@163.com

中圖分類號: TN402
計量
- 文章訪問數(shù): 182
- HTML全文瀏覽量: 93
- PDF下載量: 25
- 被引次數(shù): 0
出版歷程
- 收稿日期: 2024-03-12
- 修回日期: 2024-09-12
- 網(wǎng)絡(luò)出版日期: 2024-09-28
- 刊出日期: 2024-11-01

A Design of On-chip Network with Self-adaptive Fault-Tolerant Link

1.
School of Computer and Information , Anhui Polytechnic University, Wuhu 241000, China
2.
School of Computer Science and Information Engineering, Hefei University of Technology, Hefei 230009, China
3.
School of Microlectronics, Hefei University of Technology, Hefei 230009, China

Funds: The National Natural Science Foundation of China (62374049, 62174048, 62274052)

摘要

摘要: 隨著芯片制程不斷深入到亞微納米級別，技術(shù)節(jié)點的持續(xù)縮小加速了片上網(wǎng)絡(luò)中鏈路故障的發(fā)生。故障鏈路的增多降低了可用的路由路徑數(shù)量，并可能導(dǎo)致嚴(yán)重的流量擁塞甚至系統(tǒng)崩潰。為了保證在遭遇故障鏈路時數(shù)據(jù)包的正常傳輸，該文提出一種基于自適應(yīng)容錯鏈路的片上網(wǎng)絡(luò)設(shè)計(AFL_NoC)，它能夠?qū)⒃庥龉收湘溌返臄?shù)據(jù)包轉(zhuǎn)發(fā)到另一條可逆鏈路上。該方案包括了可逆鏈路的具體實現(xiàn)以及相應(yīng)的分布式控制協(xié)議。這種動態(tài)容錯鏈路設(shè)計充分利用了網(wǎng)絡(luò)中空閑的可用鏈路資源，確保了在遭遇鏈路故障的情況下網(wǎng)絡(luò)通信不會中斷。與先進的容錯偏轉(zhuǎn)路由算法QFCAR-W相比，AFL_NoC平均延遲降低10%，面積開銷減少了14.2%，功耗開銷減少了9.3%。
- 片上系統(tǒng) /
- 片上網(wǎng)絡(luò) /
- 容錯 /
- 路由器架構(gòu) /
- 路由算法
Abstract: As chip manufacturing has advanced to the sub-micro-nanometer scale, shrinking technology nodes are accelerating link failures in on-chip network, and the growth of failure links reduces the number of available routing paths and might lead to severe traffic congestion or even system crashes. The difficulty in maintaining the correctness of the on-chip system dramatically rises as the technology node shrinks. Previous schemes typically utilize deflection algorithms to bypass packets. However, they incur additional transmission latency due to hop count and raise the probability of deadlock. In order to achieve normal packet transmission when encountering faulty links, a self-Adaptive Fault-tolerant Link NoC design (AFL_NoC) is proposed, which redirects packets encountering a faulty link to another reversible link. The scheme contains a specific implementation of the reversible link and the associated distributed control protocol. The dynamic fault-tolerant link design fully utilizes the idle, available link and ensures that the network communication is not interrupted in case of link failures. Compared with the advanced fault-tolerant deflection routing algorithm QFCAR-W, AFL_NoC can reduce the average delay by 10%, the area overhead by 14.2%, and the power overhead by 9.3%.
- System-on-Chip (SoC) /
- Network-on-Chip (NoC) /
- Fault-tolerance /
- Router architecture /
- Routing algorithm

HTML全文

圖 1 可逆鏈路功能與電路實現(xiàn)

下載: 全尺寸圖片幻燈片

圖 2 可逆鏈路的兩個功能實現(xiàn)

下載: 全尺寸圖片幻燈片

圖 3 RFL_NoC架構(gòu)

下載: 全尺寸圖片幻燈片

圖 4 路由器間自適應(yīng)鏈路的控制機制

下載: 全尺寸圖片幻燈片

圖 5 路由器間鏈路的主/從狀態(tài)機

下載: 全尺寸圖片幻燈片

圖 6 路由器間鏈路的永久故障

下載: 全尺寸圖片幻燈片

圖 7 面向鏈路故障的容錯設(shè)計

下載: 全尺寸圖片幻燈片

圖 8 不同流量模式下，各個方案在不同故障率的平均延遲隨網(wǎng)絡(luò)注入率的變化

下載: 全尺寸圖片幻燈片

圖 9 不同流量模式下，各個方案在不同鏈路故障率的飽和吞吐量

下載: 全尺寸圖片幻燈片

參考文獻(20)

[1]	LIANG Huaguo, XU Xiumin, HUANG Zhengfeng, et al. A methodology for characterization of SET propagation in SRAM-based FPGAs[J]. IEEE Transactions on Nuclear Science, 2016, 63(6): 2985–2992. doi: 10.1109/TNS.2016.2620165.
[2]	WANG Ke and LOURI A. CURE: A high-performance, low-power, and reliable network-on-chip design using reinforcement learning[J]. IEEE Transactions on Parallel and Distributed Systems, 2020, 31(9): 2125–2138. doi: 10.1109/TPDS.2020.2986297.
[3]	BHOWMIK B. Maximal connectivity test with channel-open faults in on-chip communication networks[J]. Journal of Electronic Testing, 2020, 36(3): 385–408. doi: 10.1007/S10836-020-05878-1.
[4]	DITOMASO D, BORATEN T, KODI A, et al. Dynamic error mitigation in NoCs using intelligent prediction techniques[C]. 2016 49th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO), Taipei, China, 2016: 1–12. doi: 10.1109/MICRO.2016.7783734.
[5]	CHANG Y C, GONG C S A, and CHIU C T. Fault-tolerant mesh-based NoC with router-level redundancy[J]. Journal of Signal Processing Systems, 2020, 92(4): 345–355. doi: 10.1007/S11265-019-01476-3.
[6]	GUO Pengxing, HOU Weigang, GUO Lei, et al. Fault-tolerant routing mechanism in 3D optical network-on-chip based on node reuse[J]. IEEE Transactions on Parallel and Distributed Systems, 2020, 31(3): 547–564. doi: 10.1109/TPDS.2019.2939240.
[7]	NARAYANASAMY P and GOPALAKRISHNAN S. Novel fault tolerance topology using corvus seek algorithm for application specific NoC[J]. Integration, 2023, 89: 146–154. doi: 10.1016/J.VLSI.2022.11.011.
[8]	SLEEBA S Z, JOSE J, and MINI M G. Energy-efficient fault tolerant technique for deflection routers in two-dimensional mesh Network-on-Chips[J]. IET Computers & Digital Techniques, 2018, 12(3): 69–79. doi: 10.1049/IET-CDT.2017.0006.
[9]	SAMALA J, TAKAWALE H, CHOKHANI Y, et al. Fault-tolerant routing algorithm for mesh based NoC using reinforcement learning[C]. 2020 24th International Symposium on VLSI Design and Test (VDAT), Bhubaneswar, India, 2020: 1–6. doi: 10.1109/VDAT50263.2020.9190340.
[10]	LIU Yi, GUO Rujia, XU Changqing, et al. A Q-learning-based fault-tolerant and congestion-aware adaptive routing algorithm for networks-on-chip[J]. IEEE Embedded Systems Letters, 2022, 14(4): 203–206. doi: 10.1109/LES.2022.3176233.
[11]	JAIN A, LAXMI V, TRIPATHI M, et al. TRACK: An algorithm for fault-Tolerant, dynamic and scalable 2D mesh network-on-chip routing reconfiguration[J]. Integration, 2020, 72: 92–110. doi: 10.1016/J.VLSI.2020.01.005.
[12]	ZHANG Ying, HONG Xinpeng, CHEN Zhongsheng, et al. A deterministic-path routing algorithm for tolerating many faults on very-large-scale network-on-chip[J]. ACM Transactions on Design Automation of Electronic Systems (TODAES), 2021, 26(1): 8. doi: 10.1145/3414060.
[13]	LI Jiao, QIN Chaoqun, and SUN Xuecheng. An efficient adaptive routing algorithm for the Co-optimization of fault tolerance and congestion awareness based on 3D NoC[J]. Microelectronics Journal, 2023, 142: 105989. doi: 10.1016/J.MEJO.2023.105989.
[14]	RIZK M, MARTIN K J M, and DIGUET J P. Run-time remapping algorithm of dataflow actors on NoC-based heterogeneous MPSoCs[J]. IEEE Transactions on Parallel and Distributed Systems, 2022, 33(12): 3959–3976. doi: 10.1109/TPDS.2022.3177957.
[15]	WANG K, LOURI A, KARANTH A, et al. IntelliNoC: A holistic design framework for energy-efficient and reliable on-chip communication for manycores[C]. Proceedings of the 46th International Symposium on Computer Architecture, Phoenix, USA, 2019: 589–600. doi: 10.1145/3307650.3322274.
[16]	ZHENG Hao and LOURI A. Agile: A learning-enabled power and performance-efficient network-on-chip design[J]. IEEE Transactions on Emerging Topics in Computing, 2022, 10(1): 223–236. doi: 10.1109/TETC.2020.3003496.
[17]	LAN Y C, LIN H A, LO S H, et al. A bidirectional NoC (BiNoC) architecture with dynamic self-reconfigurable channel[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2011, 30(3): 427–440. doi: 10.1109/TCAD.2010.2086930.
[18]	FARROKHBAKHT H, KAO H, HASAN K, et al. Pitstop: Enabling a virtual network free network-on-chip[C]. 2021 IEEE International Symposium on High-Performance Computer Architecture (HPCA), Seoul, Korea (South), 2021: 682–695. doi: 10.1109/HPCA51647.2021.00063.
[19]	SUN Chen, CHEN C H O, KURIAN G, et al. DSENT-a tool connecting emerging photonics with electronics for opto-electronic networks-on-chip modeling[C]. 2012 IEEE/ACM Sixth International Symposium on Networks-on-Chip, Lyngby, Denmark, 2012: 201–210. doi: 10.1109/NOCS.2012.31.
[20]	ZHOU Wu, OUYANG Yiming, XU Dongyu, et al. Energy-efficient multiple network-on-chip architecture with bandwidth expansion[J]. IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2023, 31(4): 442–455. doi: 10.1109/TVLSI.2023.3244859.