Phase Change Random Access Memory for Neuro‐Inspired Computing

Abstract

Neuro‐inspired computing using emerging memristors plays an increasingly significant role for the realization of artificial intelligence and thus has attracted widespread interest in the era of big data. Thanks to the maturity of technology and the superiority of device performance, phase change random access memory (PCRAM) is a promising candidate for both nonvolatile memories and neuro‐inspired computing. Recently many efforts have been carried out to achieve the biological behavior using PCRAM and to clarify the related working mechanism. In order to further improve device performances, it is helpful and urgent to summarize and discuss the PCRAM solution for neuro‐inspired computing. In this paper, fundamentals, principles, recent progresses, existing challenges, and mainstream solutions are reviewed, and a brief outlook is highlighted and introduced, with the expectation to expound future directions.

Realization of volatile and non-volatile resistive switching with N-TiO2 nanorod arrays based memristive devices through compositional control

“…For a synapse, the Ca 2+ ion concentration increases when a membrane action potential is firing. The Ca 2+ ions will return to their original level after withdrawing the firing pulse [42]. The decay feature in the three memristive devices are consistent to the synaptic relaxation.…”

Dot‐Product Operation in Crossbar Array Using a Self‐Rectifying Resistive Device

“…Various non-volatile memory devices, such as ferroelectric random access memory (FRAM), phase-change random access memory (PRAM), spin-torquetransfer magnetic random access memory (STT-MRAM), and resistive switching random access memory (RRAM) have been considered for use as a memory element. [11][12][13][14][15][16][17][18][19][20][21][22] Among these devices, the RRAM exhibits outstanding characteristics, such as device scaling down, low power consumption, fast operation speed, simple structure, and high reliability.Various selection devices (SD) such as diodes, [23,24] metalinsulator-transition (MIT) devices, [25,26] ovonic threshold switch (OTS), [27][28][29] mixed-ionic-electronic-conductor (MIEC), [30,31] tunneling-oxide-based devices, [32,33] and timing selector [34] have been proposed to add the selection function to the RRAM devices. Furthermore, a 1-transistor and 1-resistor (1T1R) has been proposed as an active unit cell.…”

“…Various non-volatile memory devices, such as ferroelectric random access memory (FRAM), phase-change random access memory (PRAM), spin-torquetransfer magnetic random access memory (STT-MRAM), and resistive switching random access memory (RRAM) have been considered for use as a memory element. [11][12][13][14][15][16][17][18][19][20][21][22] Among these devices, the RRAM exhibits outstanding characteristics, such as device scaling down, low power consumption, fast operation speed, simple structure, and high reliability.…”

Interfacial Ion‐Trapping Electrolyte‐Gated Transistors for High‐Fidelity Neuromorphic Computing

“…In particular, long‐term synaptic plasticity is the essential key to enable efficient neuromorphic computing with the function of non‐volatile memory. [ 4–6 ] To mimic the synaptic functions in neuromorphic computing, so‐called ‘synaptic devices’ based on two‐terminals, such as resistive random access memory (ReRAM), [ 7–9 ] ferroelectric random access memory (FeRAM), [ 10 ] and phase change memory (PCRAM), [ 11 ] have been introduced. Linearly and symmetrically updating the electrical conductance across the two terminals, however, is very challenging as the two‐terminal devices have to write and read memory via the same terminals.…”