RescueSNN：在永久性故障下实现脉冲神经网络加速器上的可靠执行
RescueSNN: Enabling Reliable Executions on Spiking Neural Network Accelerators under Permanent Faults
AI总结：
针对SNN芯片永久性故障影响精度的问题，提出无需重训练的RescueSNN方法，通过故障感知映射和轻量级硬件增强，在高故障率下将精度提升高达80%且吞吐量降低不超过25%。
摘要：
为了最大化资源受限嵌入式系统上脉冲神经网络（SNN）处理的性能和能效，人们采用了专门的硬件加速器/芯片。然而，这些SNN芯片可能会遭受永久性故障，从而影响权重存储器和神经元行为的功能，进而可能导致显著的精度下降和系统故障。这种永久性故障可能来自制造过程中的制造缺陷，和/或运行期间由于磨损导致的器件/晶体管损坏。然而，SNN芯片中永久性故障的影响及其相应的缓解技术尚未得到彻底研究。为此，我们提出了RescueSNN，这是一种在不要求额外重新训练的情况下缓解SNN芯片计算引擎中永久性故障的新颖方法，从而在保持吞吐量和质量的同时，显著削减了设计时间和重新训练成本。我们RescueSNN方法的核心思想是：（1）分析SNN在永久性故障下的特性；（2）利用该分析通过有效的故障感知映射（FAM）来提高SNN的容错能力；以及（3）设计轻量级硬件增强以支持FAM。我们的FAM技术利用SNN计算引擎的故障图来（i）在将权重位映射到故障存储单元时最小化权重损坏，以及（ii）在考虑SNN操作和处理数据流的同时，选择性地使用不会导致显著精度下降的故障神经元以维持精度和吞吐量。实验结果表明，与在没有缓解措施的情况下在故障芯片上运行SNN相比，我们的RescueSNN在高故障率（例如，潜在故障位置的0.5）下，将精度提高了高达80%，同时将吞吐量降低维持在25%以下。
英文摘要：
To maximize the performance and energy efficiency of Spiking Neural Network (SNN) processing on resource-constrained embedded systems, specialized hardware accelerators/chips are employed. However, these SNN chips may suffer from permanent faults which can affect the functionality of weight memory and neuron behavior, thereby causing potentially significant accuracy degradation and system malfunctioning. Such permanent faults may come from manufacturing defects during the fabrication process, and/or from device/transistor damages (e.g., due to wear out) during the run-time operation. However, the impact of permanent faults in SNN chips and the respective mitigation techniques have not been thoroughly investigated yet. Toward this, we propose RescueSNN, a novel methodology to mitigate permanent faults in the compute engine of SNN chips without requiring additional retraining, thereby significantly cutting down the design time and retraining costs, while maintaining the throughput and quality. The key ideas of our RescueSNN methodology are (1) analyzing the characteristics of SNN under permanent faults; (2) leveraging this analysis to improve the SNN fault-tolerance through effective fault-aware mapping (FAM); and (3) devising lightweight hardware enhancements to support FAM. Our FAM technique leverages the fault map of SNN compute engine for (i) minimizing weight corruption when mapping weight bits on the faulty memory cells, and (ii) selectively employing faulty neurons that do not cause significant accuracy degradation to maintain accuracy and throughput, while considering the SNN operations and processing dataflow. The experimental results show that our RescueSNN improves accuracy by up to 80% while maintaining the throughput reduction below 25% in high fault rate (e.g., 0.5 of the potential fault locations), as compared to running SNNs on the faulty chip without mitigation.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
