Toward memristive in-memory computing: principles and applications

Abstract

With the rapid growth of computer science and big data, the traditional von Neumann architecture suffers the aggravating data communication costs due to the separated structure of the processing units and memories. Memristive in-memory computing paradigm is considered as a prominent candidate to address these issues, and plentiful applications have been demonstrated and verified. These applications can be broadly categorized into two major types: soft computing that can tolerant uncertain and imprecise results, and hard computing that emphasizes explicit and precise numerical results for each task, leading to different requirements on the computational accuracies and the corresponding hardware solutions. In this review, we conduct a thorough survey of the recent advances of memristive in-memory computing applications, both on the soft computing type that focuses on artificial neural networks and other machine learning algorithms, and the hard computing type that includes scientific computing and digital image processing. At the end of the review, we discuss the remaining challenges and future opportunities of memristive in-memory computing in the incoming Artificial Intelligence of Things era. Graphical Abstract

Emerging MXene‐Based Memristors for In‐Memory, Neuromorphic Computing, and Logic Operation

“…[12,13] Recently, memristors, which were theoretically proposed by Leon Chua [14] and first demonstrated by Hewlett Packard laboratory, [15] have aroused tremendous attention due to the advantages of excellent data storage capability, low operating energy, fast response, and in-memory computing potential. [16][17][18][19][20] Different from the traditional Flash memory and random-access memory that rely on charges, memristors can alter their resistances to the history of electrical input, offering the media to store information in the form of resistance switching. [21][22][23] Moreover, memristors with reconfigurable history-dependent resistive switching behaviors show great promise to mimic biological synapses, which emerge as one of the most fascinating technologies for implementing artificial neural network for neuromorphic computing.…”

Low-dimensional nanostructures for monolithic 3D-integrated flexible and stretchable electronics

“…This necessitates novel circuit-building structures to overcome the declining costeffectiveness of transistor scaling and the inherent inefficiency of using transistors in non-von Neumann computing architecture. 248,249 A memristor is an emerging electronic device that can switch between resistance states, retaining charges without power. It provides non-volatile storage, high endurance, fast speed, high-density integration, and ultra-low power dissipation.…”

Nanowire-based synaptic devices for neuromorphic computing

“…In addition to pattern recognition, neuromorphic computing can also be used for image processing, speech recognition and motion monitoring [78]. However, NW synaptic devices have been rarely studied for applications in these areas.…”