Exploring Phase‐Change Memory: From Material Systems to Device Physics

Ren, Yumei; Sun, Ruoyao; Chen, Stephenie Hiu Yuet; Du, Chunyu; Han, Su‐Ting; Zhou, Ye

doi:10.1002/pssr.202000394

Cited by 11 publications

(9 citation statements)

References 155 publications

(228 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Consequently, the development of novel devices is paramount for the advancement of neuromorphic computing. At present, various devices have been developed to realize brain-inspired computing, represented by resistive switching devices and neuromorphic transistors [14][15][16][17][18]. In terms of device functions, neurons, and synaptic behaviors have been successfully simulated.…”

Section: Introductionmentioning

confidence: 99%

CMOS-compatible neuromorphic devices for neuromorphic perception and computing: a review

Zhu,

Mao,

Zhu

et al. 2023

Int. J. Extrem. Manuf.

View full text Add to dashboard Cite

Neuromorphic computing is a brain-inspired computing paradigm that aims to construct efficient, low-power, and adaptive computing systems by emulating the information processing mechanisms of biological neural systems. At the core of neuromorphic computing are neuromorphic devices that mimic the functions and dynamics of neurons and synapses, enabling the hardware implementation of artificial neural networks. Various types of neuromorphic devices have been proposed based on different physical mechanisms such as resistive switching device and electric-double-layer transistors. These devices have demonstrated a range of neuromorphic functions such as multistate storage, spike-timing-dependent plasticity, and dynamic filtering, etc. To achieve high performance neuromorphic computing systems, it is essential to fabricate neuromorphic devices compatible with complementary metal oxide semiconductor (CMOS) manufacturing process. This improves the device reliability and stability and is favourable to achieve neuromorphic chips with higher integration density and low power consumption. This review summarizes CMOS-compatible neuromorphic devices and discusses their emulation of synaptic and neuronal functions as well as their applications in neuromorphic perception and computing. We highlight challenges and opportunities for further development of CMOS-compatible neuromorphic devices and systems.

show abstract

Section: Introductionmentioning

confidence: 99%

CMOS-compatible neuromorphic devices for neuromorphic perception and computing: a review

Zhu,

Mao,

Zhu

et al. 2023

Int. J. Extrem. Manuf.

View full text Add to dashboard Cite

show abstract

“…[9,10] At the same time, aggressive requirements on processing energy and real-time response require a new computing paradigm with possible new devices and architecture. [11][12][13][14] Therefore, researchers are exploring nextgeneration memory computing devices with nonvolatile memory behavior, like ferroelectric resistive memory devices, [15][16][17] phase-change resistive memory devices, [18][19][20] and magnetic resistive memory devices. [21][22][23] Later in the 1970s, Leon Chua introduced a new memory element called memristor and considered a fourth fundamental passive circuit element.…”

Section: Introductionmentioning

confidence: 99%

Advancement in Soft Iontronic Resistive Memory Devices and Their Application for Neuromorphic Computing

Khan

Kim

Chougale

et al. 2022

Advanced Intelligent Systems

View full text Add to dashboard Cite

The aqueous electrolyte can be a deformable and stretchable liquid material for iontronic resistive memory devices. An aqueous medium makes a device closer to the brain‐like system with the movement of ions. This review paper proposes advances in liquid resistive memories and neuromorphic computing behavior to emulate electronic synapses. Primarily, the aqueous iontronic resistive memories can be used to study electrode and active layer materials and different device structures. Hence, herein, a timely and comprehensive study of these devices using ionic liquids, hydrogels, salt solutions, and soft electrodes to classify the device mechanism is presented. The filament formation is discussed in detail based on ion concentration polarization, electrode metallization, and movements of ions and charged molecules, which result in the formation of the metal dendrite. To manufacture a higher‐performance memory, device parameters should be optimized based on aqueous electrolytes, electrode materials, and other device design parameters. Aqueous electrolytes have smooth neurotransmission ability to fabricate brain‐inspired resistive memories with stable performance and device repeatability with smooth ion transmission. Aqueous electrode materials can be reliable for neural interface activities to compute electronic synapsis with electrical and chemical properties to ensure device reliability for a longer time period.

show abstract

“…In this framework, several neuromorphic spiking neural networks (SNNs) based on CMOS technology have been proposed, demonstrating VLSI synaptic circuits with homeostatic neurons (Bartolozzi and Indiveri, 2006 ; Chicca et al, 2014 ; Qiao et al, 2017 ) and reward-based decision-making circuits (Wunderlich et al, 2019 ; Yan et al, 2019 ). At the same time, non-volatile memory devices, such as phase change memory (PCM), have raised considerable interest as promising synaptic connections for neuromorphic computation, thanks to the 3D stacking capability, the low-voltage operation and the ability to serve as embedded non-volatile memory in computing systems (Suri et al, 2012 ; Xu et al, 2020 ; Ren et al, 2021 ). In particular, PCMs have recently demonstrated outstanding multi-level capability (Kuzum et al, 2013 ; Ren et al, 2021 ), which enables continual learning in neural networks (Bianchi et al, 2019 ; Muñoz-Martín et al, 2019 ) and decision making in brain-inspired cognitive systems (Eryilmaz et al, 2014 ).…”

Section: Introductionmentioning

confidence: 99%

“…At the same time, non-volatile memory devices, such as phase change memory (PCM), have raised considerable interest as promising synaptic connections for neuromorphic computation, thanks to the 3D stacking capability, the low-voltage operation and the ability to serve as embedded non-volatile memory in computing systems (Suri et al, 2012 ; Xu et al, 2020 ; Ren et al, 2021 ). In particular, PCMs have recently demonstrated outstanding multi-level capability (Kuzum et al, 2013 ; Ren et al, 2021 ), which enables continual learning in neural networks (Bianchi et al, 2019 ; Muñoz-Martín et al, 2019 ) and decision making in brain-inspired cognitive systems (Eryilmaz et al, 2014 ).…”

Section: Introductionmentioning

confidence: 99%

“…We also analyze the impact of device programming failure in relation to the multilevel capability of the PCM devices. The impact of PCM conductance drift is also studied (Suri et al, 2012 ; Xu et al, 2020 ; Ren et al, 2021 ), demonstrating that this device non-ideality could implement bio-inspired features, such as active forgetting. Finally, we propose a novel bio-inspired recurrent neural network (RNN) capable of solving reinforcement learning tasks.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Brain-Inspired Homeostatic Neuron Based on Phase-Change Memories for Efficient Neuromorphic Computing

et al. 2021

View full text Add to dashboard Cite

One of the main goals of neuromorphic computing is the implementation and design of systems capable of dynamic evolution with respect to their own experience. In biology, synaptic scaling is the homeostatic mechanism which controls the frequency of neural spikes within stable boundaries for improved learning activity. To introduce such control mechanism in a hardware spiking neural network (SNN), we present here a novel artificial neuron based on phase change memory (PCM) devices capable of internal regulation via homeostatic and plastic phenomena. We experimentally show that this mechanism increases the robustness of the system thus optimizing the multi-pattern learning under spike-timing-dependent plasticity (STDP). It also improves the continual learning capability of hybrid supervised-unsupervised convolutional neural networks (CNNs), in terms of both resilience and accuracy. Furthermore, the use of neurons capable of self-regulating their fire responsivity as a function of the PCM internal state enables the design of dynamic networks. In this scenario, we propose to use the PCM-based neurons to design bio-inspired recurrent networks for autonomous decision making in navigation tasks. The agent relies on neuronal spike-frequency adaptation (SFA) to explore the environment via penalties and rewards. Finally, we show that the conductance drift of the PCM devices, contrarily to the applications in neural network accelerators, can improve the overall energy efficiency of neuromorphic computing by implementing bio-plausible active forgetting.

show abstract

Exploring Phase‐Change Memory: From Material Systems to Device Physics

Cited by 11 publications

References 155 publications

CMOS-compatible neuromorphic devices for neuromorphic perception and computing: a review

CMOS-compatible neuromorphic devices for neuromorphic perception and computing: a review

Advancement in Soft Iontronic Resistive Memory Devices and Their Application for Neuromorphic Computing

A Brain-Inspired Homeostatic Neuron Based on Phase-Change Memories for Efficient Neuromorphic Computing

Contact Info

Product

Resources

About