arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2402.10981cs.ARcs.LGcs.NE
收藏

ReRAM神经形态电路阵列中的固定故障及其机器学习校正方法

Stuck-at Faults in ReRAM Neuromorphic Circuit Array and their Correction through Machine Learning

San Jose State University(圣何塞州立大学)

Vedant Sawal, Hiu Yung Wong

AI总结:

针对ReRAM神经形态电路的固定故障导致推理精度下降的问题,研究了不同缺陷模式的影响,并提出机器学习校正策略,将缺陷电路推理精度从48%提升至85%。

摘要:

本文研究了阻变随机存取存储器(Resistive Random Access Memory,ReRAM)神经形态电路因固定故障(恒通、恒断以及固定在某一电阻值)导致的推理精度问题。研究采用基于Python的仿真框架,对手写数字数据集开展有监督机器学习(神经网络包含3个隐藏层、1个输入层和1个输出层),并构建对应的由Spectre仿真的全模拟神经形态电路(含4个突触阵列)。仿真采用通用的45nm工艺开发套件(Process Development Kit,PDK)。研究分析了恒通和恒断缺陷导致的推理精度下降差异,探究了圆形、环形、行、列以及圆形补集等多种缺陷模式。研究发现,恒通和恒断缺陷对推理精度的影响相近;但若缺陷在列方向存在空间分布差异,推理精度可能会显著下降。本文还提出了一种机器学习(ML)策略,用于恢复固定故障导致的推理精度下降,可将存在缺陷的神经形态电路的推理精度从48%提升至85%。

英文摘要:

In this paper, we study the inference accuracy of the Resistive Random Access Memory (ReRAM) neuromorphic circuit due to stuck-at faults (stuck-on, stuck-off, and stuck at a certain resistive value). A simulation framework using Python is used to perform supervised machine learning (neural network with 3 hidden layers, 1 input layer, and 1 output layer) of handwritten digits and construct a corresponding fully analog neuromorphic circuit (4 synaptic arrays) simulated by Spectre. A generic 45nm Process Development Kit (PDK) was used. We study the difference in the inference accuracy degradation due to stuck-on and stuck-off defects. Various defect patterns are studied including circular, ring, row, column, and circular-complement defects. It is found that stuck-on and stuck-off defects have a similar effect on inference accuracy. However, it is also found that if there is a spatial defect variation across the columns, the inference accuracy may be degraded significantly. We also propose a machine learning (ML) strategy to recover the inference accuracy degradation due to stuck-at faults. The inference accuracy is improved from 48% to 85% in a defective neuromorphic circuit.

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。

↑