Essential Characteristics of Memristors for Neuromorphic Computing (Adv. Electron. Mater. 2/2023)

Abstract

Memristors Memristive neuromorphic computing promotes the development of humanoid robotics, and its performance is directly affected by the memristors' characteristics. In article number 2200833, Wenbin Chen, Shuo Gao, and co‐workers overview the recent progress of memristors with different physical mechanisms and their characteristics. The existing issues and challenges in implementing neuromorphic computing systems based on these different mechanisms are highlighted and prospected.

Reliable Memristor Crossbar Array Based on 2D Layered Nickel Phosphorus Trisulfide for Energy‐Efficient Neuromorphic Hardware

“…Long‐term plasticity (LTPL) characteristics including long‐term potentiation (LTP) and long‐term depression (LTD) are the most important characteristic of the biological synapse and the basic requirement for electronic synapses in ANNs. [ 45 ] As illustrated in Figure a, more than 55 pulses were required for the evolution from short‐term potentiation (STP) to LTP when the memristor operated in total‐reset mode. In contrast, the memristor operating in quasi‐reset mode did not exhibit this transition behavior (Figure 3b), indicating that the memristor operating in quasi‐reset mode can mimic long‐term plasticity better than that in total‐reset mode.…”

Wood-derived Fe cluster-reinforced asymmetric single-atom catalysts and weather-resistant organohydrogel for wide-temperature flexible Zn–air batteries

“…Consistent with the previous findings, the presence of S atoms and the proximity of the Fe cluster contributed to a reduction in electron donation from the Fe atom toward neighboring N atoms. [48][49][50] Consequently, this alteration affected both the hybridization state of Fe and adjacent N atoms, while optimizing the local electronic structure at the catalytic sites, ultimately enhancing the ORR catalytic activity. 51 The electronic structures of the Fe 4 -FeN 3 S 1 C, FeN 3 S 1 -C, Fe 4 -FeN 4 C, and FeN 4 C models were further assessed using the d-band center theory and partial density of states.…”

Analog Ion‐Slicing LiNbO3 Memristor Based on Hopping Transport for Neuromorphic Computing

“…[8,16] However, the existing analog memristors still suffer from large device and cyclic variations, which significantly decreases the precision of weight updating. [9,17,18] For example, the migration of oxygen vacancies (V o s) or metal ions in a memristor based on amorphous switching oxides is stochastic due to a large number of disorders and defect clusters in amorphous oxides. [10,19,20] Unfortunately, these intrinsically random defects are formed unavoidably during common oxide film growing processes, such as layer deposition and magnetron sputtering.…”