Recent Progress on Memristive Convolutional Neural Networks for Edge Intelligence

• Published in: Advanced intelligent systems 2020-11, Vol.2 (11), p.n/a
• Main Authors: Qin, Yi-Fan, Bao, Han, Wang, Feng, Chen, Jia, Li, Yi, Miao, Xiang-Shui
• Format: Article
• Language: English
• Subjects: Accelerators; Arrays; Artificial intelligence; Artificial neural networks; Bandwidths; Cloud computing; CMOS; Computer memory; convolutional neural network accelerators; Deep learning; edge intelligence; Energy consumption; Energy efficiency; Image compression; Image processing; Information storage; Internet of Things; long short-term memory neural networks; Memristors; Neural networks; Optimization; quantization; Random access memory; Semiconductors; Transistors
• ISSN: 2640-4567; 2640-4567
• DOI: 10.1002/aisy.202000114

Recently, due to the development of big data and computer technology, artificial intelligence (AI) has received extensive attention and made great progress. Edge intelligence pushes the computing center of AI from the cloud to individual users, making AI closer to life, but at the same time puts forward higher requirements for the realization of hardware, especially for edge acceleration. Taking convolutional neural networks (CNNs) as an example, which show excellent problem‐solving capabilities in different fields of academia and industry, it still faces issues of enormous computing volume and complex mapping architecture. Based on the computing‐in‐memory property and parallel multiply accumulate (MAC) operations of the emerging nonvolatile memristor arrays, herein the recent research progress of the edge intelligence memristive convolution accelerator is summarized. Furthermore, aiming at improving memristive convolutional accelerators, two potential optimization schemes are also discussed: The compression methods represented by quantization show great potential for static image processing, and the combination of a CNN with a long short‐term memory (LSTM) neural network makes up for the CNN's shortcomings of dynamic target processing. Finally, the future challenges and opportunities of edge intelligence accelerators based on memristor arrays are also discussed. Convolutional neural network (CNN) accelerators based on memristive crossbars are promising for hardware realization of edge intelligence. Herein, the recent advances in memristor‐driven CNN acceleration are introduced, mainly focusing on the analog/quantized networks for static image processing and the combination of a long short‐term memory (LSTM) neural network and a CNN for dynamic tasks.