Fine-grained fault-tolerant adaptive routing for networks-on-chip

Liu, J; Harkin, J; Maguire, L; Li, Y; Wan, L; Luo, Y

Abstract

Due to the increase of physical defects in advanced manufacturing processes, Networks-on-Chip (NoC) system reliability is a critical challenge as faults often occur post manufacturing. Therefore it is important to add fault tolerance to the NoC system. In this paper, a novel routing algorithm for 2D mesh NoCs is proposed which aims to enhance the fault-tolerant capabilities via a look-ahead function. A traffic status informing mechanism is developed to provide information to local NoC routers on the interconnect conditions in far distant routers. In addition, a weighted path mechanism is used to forward the packets. The routing algorithm is implemented and verified on FPGA hardware. Real-time throughput and traffic information were collected by a monitoring unit on the FPGA. Results show that the proposed routing algorithm can maintain the system function under low fault rates and only has a marginal (~5 %) throughput degradation under high fault rate of 20 %. The router area is also relatively low which demonstrated its scalability.

Citation

Liu, J., Harkin, J., Maguire, L., Li, Y., Wan, L., & Luo, Y. (2015). Fine-grained fault-tolerant adaptive routing for networks-on-chip. In Algorithms and Architectures for Parallel Processing : 15th International Conference, ICA3PP 2015, Zhangjiajie, China, November 18-20, 2015, Proceedings, Part IV (492-505). Springer International Publishing. https://doi.org/10.1007/978-3-319-27140-8_34

• Publication Date: Dec 16, 2015
• Deposit Date: Jan 6, 2016
• Pages: 492-505
• Series Title: Lecture Notes in Computer Science
• Series Number: 9531
• Book Title: Algorithms and Architectures for Parallel Processing : 15th International Conference, ICA3PP 2015, Zhangjiajie, China, November 18-20, 2015, Proceedings, Part IV
• ISBN: 9783319271408
• DOI: https://doi.org/10.1007/978-3-319-27140-8_34
• Publisher URL: http://dx.doi.org/10.1007/978-3-319-27140-8_34
• Related Public URLs: http://dx.doi.org/10.1007/978-3-319-27140-8