Communication Synchronization-Aware Arbitration Policy in NoC-Based DNN Accelerators

Wenjie Fan, Siyue Li, Lingxiao Zhu, Zhonghai Lu, Yuxiang FuIEEE Transactions on Circuits & Systems II Express Briefs 2024 · DOI

Abstract

In NoC-based neural network accelerators, many-to-one and many-to-many are prevalent traffic patterns. In these traffic patterns, there exists a need for communication synchronization between Processing Elements (PEs) of adjacent layers to optimize latency. The last received packet will determine the end time of a layer’s computation. The Communication Synchronization-aware Arbitration Policy (CSAP) is proposed in this paper to handle this problem, which uses a negative feedback mechanism to regulate the packet sending rate of each source node. Compared with the local-age-based policy, CSAP decreases the execution by 4.69% ~ 12.55% across neural networks of different scales. The proposed policy only incurs 1.06% additional hardware overhead in the router compared with the local-age-based policy.
BibTeX
@article{fan2024communication,
  author = {Wenjie Fan and Siyue Li and Lingxiao Zhu and Zhonghai Lu and Li Li and Yuxiang Fu},
  title = {{Communication Synchronization-Aware Arbitration Policy in NoC-Based DNN Accelerators}},
  journal = {IEEE Transactions on Circuits \& Systems II Express Briefs},
  year = {2024},
  doi = {10.1109/tcsii.2024.3395054},
}

Loading related content...

Figure