Influence of variability on the performance of HfO2 memristor-based convolutional neural networks. (November 2021)

• Record Type: Journal Article
• Authors: Romero-Zaliz, R.; Pérez, E.; Jiménez-Molinos, F.; Wenger, C.; Roldán, J.B.
• Is Part Of: Solid-state electronics. Volume 185(2021)
• Publication Date: 2021-11
• Subjects: Memristors -- Multilevel RRAMs -- Hardware neural networks -- Variability; Semiconductors -- Periodicals; Semiconducteurs -- Périodiques; 621.38152
• DOI: 10.1016/j.sse.2021.108064
• Languages: English
• ISSNs: 0038-1101

Abstract

Highlights: Quantization and variability in neural network synaptic weights was studied. Different convolutional neural network architectures were considered. A multilevel approach for HfO2 -based memristors was employed for quantization. Abstract: A study of convolutional neural networks (CNNs) was performed to analyze the influence of quantization and variability in the network synaptic weights. Different CNNs were considered accounting for the number of convolutional layers, size of the filters in the convolutional layer, number of neurons in the final network layers and different sets of quantization levels. The conductance levels of fabricated 1T1R structures based on HfO2 memristors were considered as reference for four or eight level quantization processes at the inference stage of the CNNs, which were previous trained with the MNIST dataset. We also included the variability of the experimental conductance levels that was found to be Gaussian distributed and was correspondingly modeled for the synaptic weight implementation.