Stuck-at Faults in ReRAM Neuromorphic Circuit Array and their Correction through Machine Learning
Publication Date
1-1-2024
Document Type
Conference Proceeding
Publication Title
2024 IEEE Latin American Electron Devices Conference, LAEDC 2024
DOI
10.1109/LAEDC61552.2024.10555838
Abstract
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 stuckon 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.
Keywords
Error-Correction, Inference Accuracy, Neuromorphic, ReRAM, SPICE simulation, Stuck-at Faults
Department
Electrical Engineering
Recommended Citation
Vedant Sawal and Hiu Yung Wong. "Stuck-at Faults in ReRAM Neuromorphic Circuit Array and their Correction through Machine Learning" 2024 IEEE Latin American Electron Devices Conference, LAEDC 2024 (2024). https://doi.org/10.1109/LAEDC61552.2024.10555838
