片上网络中一种改进的非暂态故障检测方法

An Improved Detection Method of Permanent Faults in On-chip Networks

摘要:

提出一种覆盖范围广、成本低的片上网络非暂态故障在线检测和诊断系统；通过使用差错症状采集和报文/数据块计数技术,实现了片上网络数据路径和控制逻辑的高分辨率故障诊断；仿真实验结果表明,该系统既可处理数据路径故障也可处理控制逻辑故障,面积开销很低(低于10%),且无需注入任何测试数据流,不会造成严重的性能开销；如果网络相对健康,则系统性能开销可以忽略不计。

Abstract:

This paper proposes a comprehensive yet low-cost solution for online detection and diagnosis of permanent faults in on-chip networks. Using error syndrome collection and packet/flit-counting techniques, high resolution defect diagnosis is feasible in both data path and control logic of the on-chip network. The simulation results show that, the proposed method addresses both data path and control-logic faults and incurs a low area overhead (less than 10%), and without injecting any test traffic or incurring significant performance overhead. For a relatively healthy network, the performance overhead of our method is negligible.