AIDX: Adaptive Inference Scheme to Mitigate State-Drift in Memristive VMM Accelerators

2020/09/01 by Tony Liu, Liu, Tony, Amirali Amirsoleimani +9
Engineering · Neuroscience · #Advanced Memory and Neural Computing #CCD and CMOS Imaging Sensors #Emerging Technologies (cs.ET) #FOS: Computer and information sciences #Photoreceptor and optogenetics research

paper · pdf · doi:10.48550/arxiv.2009.00180

openalex publication_date 2020/09/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

An adaptive inference method for crossbar (AIDX) is presented based on an optimization scheme for adjusting the duration and amplitude of input voltage pulses. AIDX minimizes the long-term effects of memristance drift on artificial neural network accuracy. The sub-threshold behavior of memristor has been modeled and verified by comparing with fabricated device data. The proposed method has been evaluated by testing on different network structures and applications, e.g., image reconstruction and classification tasks. The results showed an average of 60% improvement in convolutional neural network (CNN) performance on CIFAR10 dataset after 10000 inference operations as well as 78.6% error reduction in image reconstruction.

Related

• PRUNIX: Non-Ideality Aware Convolutional Neural Network Pruning for Memristive Accelerators
• MemSE: Fast MSE Prediction for Noisy Memristor-Based DNN Accelerators
• Design and simulation of memristor-based neural networks
• GENIEx: A Generalized Approach to Emulating Non-Ideality in Memristive\n Xbars using Neural Networks
• Representable Matrices: Enabling High Accuracy Analog Computation for Inference of DNNs using Memristors