Yue Hao , Huaqiang Wu , Yuchao Yang , Qi Liu , Xiao Gong , Genquan Han , and Ming Li

1School of Microelectronics, Xidian University, Xi’an 710071, China

2Institute of Microelectronics, Tsinghua University, Beijing 100084, China

3Department of Micro/nanoelectronics, Peking University, Beijing 100871, China

4Frontier Institute of Chip and System, Fudan University, Shanghai 200438, China

5Department of Electrical and Computer Engineering, National University of Singapore, Singapore 117546, Singapore

6Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China

Traditional charge-based memories, such as dynamic random-access memory (DRAM) and flash, are approaching their scaling limits.A variety of resistance-based memories, such as phase-change memory (PCM), magnetic random-access memory (MRAM) and resistive random-access memory(RRAM), have been long considered for emerging memory applications thanks to their non-volatility, fast speed, low power, and compact size for potentially high-density integration.More recently, they are also extensively studied as memristors for neuromorphic computing, which has quite different requirements on their resistive switching characteristics compared to digital memory application.In the past decade,tremendous progresses, from materials and physical mechanisms to devices and neuromorphic systems, have been made to advance the field.

To capture the recent advancements and report the latest breakthroughs in this field, we organized a special issue on beyond Moore, which will be published in two consecutive issues onJournal of Semiconductors.The first one, focusing on resistive switching devices for emerging memory and neuromorphic computing, consists of five comprehensive review papers and three original research articles.Sokolovet al.provides a comprehensive review on the working mechanisms, materials, and engineering techniques of memristors for emerging memory and neuromorphic computing[1].Yanget al.survey thein situtransmission electron microscopy studies on the switching mechanisms of nonvolatile memories and also their packaging reliability[2].Besides two-terminal memristors, Huanget al.review the materials and working mechanisms of electrolyte-gated transistors as promising synaptic elements for high-performance neuromorphic computing[3].Chenet al.revisit the advances in the hardware implementations of neural network accelerators and numerical computing units using memristor crossbar arrays[4].Liaoet al.review recent developments in neuromorphic vision sensors for highly efficient image processing by mimicking neuro-biological structures and functionalities of the retina[5].In addition to those review articles, Gonget al.present an original research paper on the interesting short-term synaptic plasticity in iongel gated graphene synaptic transistors with electric double layer[6].Shiet al.report the voltage-dependent plasticity, tuning from short-term plasticity to long-term potentiation, of tungsten oxide-based memristive synapses and further demonstrates the image Boolean operations in the memristive synapse array[7].Taoet al.carry out in-depth studies on the neutron irradiation effect on the electrochemical metallization memory devices, showing excellent reliability performance against neutron irradiation[8].

We sincerely hope this special issue could provide a valuable overview and perspective on the developments of emerging memory and neuromorphic computing.We also hope the creative research articles in this special issue could inspire more research interest in this field to tackle the challenges towards future applications in memory and neuromorphic systems.We would like to thank all the authors for their outstanding contributions to this special issue.We are also grateful to the editorial and production staff ofJournal of Semiconductorsfor their assistance.