Low Power Stochastic Neurons From SiO<sub>2</sub>-Based Bilayer Conductive Bridge Memristors for Probabilistic Spiking Neural Network Applications—Part II: Modeling

Bousoulas, Panagiotis; Hadfield, Jack; Aslanidis, V.; Limberopoulos, S.; Tsoukalas, D.

doi:10.1109/ted.2022.3160140

Cited by 12 publications

(5 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Published on behalf of The Japan Society of Applied Physics by IOP Publishing Ltd will be employed within a simple RC circuit to produce spikes. 77,78) From a material point of view, currently the CBRAM-based memory devices exhibit great advantages in contrast with their VCM counterparts, due to their low power consumption and their ability to operate under a dual switching mode (i.e. threshold and bipolar).…”

Section: Emulating Various Synaptic Functionalities With Pt Nps-based...mentioning

confidence: 99%

Material design strategies for emulating neuromorphic functionalities with resistive switching memories

Bousoulas

Kitsios

Chatzinikolaou

et al. 2022

Jpn. J. Appl. Phys.

Self Cite

View full text Add to dashboard Cite

Nowadays, the huge power consumption and the inability of the conventional circuits to deal with real-time classification tasks have necessitated the devising of new electronic devices with inherent neuromorphic functionalities. Resistive switching memories arise as an ideal candidate due to their low footprint and small leakage current dissipation, while their intrinsic randomness is smoothly leveraged for implementing neuromorphic functionalities. In this review, valence change memories (VCM) or conductive bridge memories (CBRAM) for emulating neuromorphic characteristics are demonstrated. Moreover, the impact of the device structure and the incorporation of $Pt$ nanoparticles (NPs) is thoroughly investigated. Interestingly, our devices possess the ability to emulate various artificial synaptic functionalities, including paired-pulsed facilitation and paired-pulse depression, long-term plasticity and four different types of spike-dependent plasticity. Our approach provides valuable insights from a material design point of view towards the development of multifunctional synaptic elements that operate with low power consumption and exhibit biological-like behavior.

show abstract

Section: Emulating Various Synaptic Functionalities With Pt Nps-based...mentioning

confidence: 99%

Material design strategies for emulating neuromorphic functionalities with resistive switching memories

Bousoulas

Kitsios

Chatzinikolaou

et al. 2022

Jpn. J. Appl. Phys.

Self Cite

View full text Add to dashboard Cite

show abstract

“…This is the reason we have first chosen Ag in our studies with MoS 2 . 19,20 On top of that, enhanced synaptic properties were recorded, regarding the linearity of the potentiation and depression procedures, which also render our structure attractive in emulating neuromorphic functionalities. 21−23…”

Section: Introductionmentioning

confidence: 96%

“…From this study we have concluded that both materials give satisfactory results; silver-based CBRAM devices exhibited both bipolar and threshold switching behavior which is of interest for the implementation of fully memristive neural networks. This is the reason we have first chosen Ag in our studies with MoS 2 . , On top of that, enhanced synaptic properties were recorded, regarding the linearity of the potentiation and depression procedures, which also render our structure attractive in emulating neuromorphic functionalities. − …”

Section: Introductionmentioning

confidence: 99%

Demonstration of Enhanced Switching Variability and Conductance Quantization Properties in a SiO₂ Conducting Bridge Resistive Memory with Embedded Two-Dimensional MoS₂ Material

Kitsios

Bousoulas

Spithouris

et al. 2022

ACS Appl. Electron. Mater.

Self Cite

View full text Add to dashboard Cite

In this work, we explore the resistive switching behavior of a thin layer of SiO 2 with embedded two-dimensional (2D) molybdenum disulfide, MoS 2 , in a conductive bridge random access memory (CBRAM) configuration. The proposed device exhibits enhanced conductance quantization behavior, reduced variability due to the suppression of the stochastic filament formation process, and synaptic properties. The device operates under the bipolar switching mode without the application of any electroforming procedure; eight different quantized conductance states were captured during direct current (DC) operation and 10 quantized states were recorded under pulse measurements. On top of that, both improved endurance and retention properties as well as linearity of the synaptic potentiation and depression procedures were attained; the underlying origins of these effects are attributed to the control of the Ag ion diffusion barrier through the existence of the atomic sieve of MoS 2 . Our work paves the way for the development of robust memristive elements for the implementation of stable resistive switching and neuromorphic functionalities.

show abstract

“…[8] Nevertheless, many neural devices used for SNNs are the mainstream complementary metal-oxidesemiconductor (CMOS) neurons that are indispensable for realizing of sophisticated circuits, which are considered unsuitable for large-scale network implementation due to their large area and low efficiency. [9,10] As a result, it is urgently expected to develop functional neuronal devices that enable the achievements of the basic functions of spiking neurons including the integration behavior and the firing action with simple circuits. Fortunately, substantial improvements have been made in developing artificial neurons by utilizing the emerging memristive devices with volatile TS characteristics, in which the switching low resistance state (LRS) can be recovered back to its initial high resistance state (HRS) in a very short period of time without applied external reset voltage.…”

Section: Introductionmentioning

confidence: 99%

“…[ 8 ] Nevertheless, many neural devices used for SNNs are the mainstream complementary metal‐oxide‐semiconductor (CMOS) neurons that are indispensable for realizing of sophisticated circuits, which are considered unsuitable for large‐scale network implementation due to their large area and low efficiency. [ 9,10 ] As a result, it is urgently expected to develop functional neuronal devices that enable the achievements of the basic functions of spiking neurons including the integration behavior and the firing action with simple circuits.…”

Section: Introductionmentioning

confidence: 99%

Artificial Neurons Using Ag−In−Zn−S/Sericin Peptide‐Based Threshold Switching Memristors for Spiking Neural Networks

He,

Yan,

Zhang

et al. 2023

Adv Elect Materials

View full text Add to dashboard Cite

Memristive devices with threshold switching characteristics can be effectively utilized to mimic biological neurons acting as one of the key building blocks for constructing advanced hardware neural networks. In this work, the emulation of leaky integrate‐and‐fire memristive neuron is realized in one single cell with Ag/Ag−In−Zn−S/silk sericin/W architecture without the need for additional auxiliary circuits. The studied devices demonstrate excellent electrical properties, such as stably repeatable threshold switching, concentratedly low threshold voltage (≈0.4 V), and relatively small device‐to‐device variation. In addition, multiple neural features, such as leaky integrate‐and‐fire neuron functionality and strength‐modulated spike frequency characteristic, have been successfully emulated owing to the forming‐free volatile threshold switching effect. The stable volatile threshold switching behaviors and regular firing event may be attributed to the controllable metallic Ag filamentary mechanism. Furthermore, a solid accuracy of 91.44% of the pattern recognition of Modified National Institute of Standards and Technology (MNIST) data is obtained via a trained spiking neural network (SNN) based on the leaky integrate‐and‐fire behavior of sericin‐based device. These achievements shed light on the fact that employing sericin biomaterials has great application potential in advanced neuromorphic computation.

show abstract

Low Power Stochastic Neurons From SiO₂-Based Bilayer Conductive Bridge Memristors for Probabilistic Spiking Neural Network Applications—Part II: Modeling

Cited by 12 publications

References 34 publications

Material design strategies for emulating neuromorphic functionalities with resistive switching memories

Material design strategies for emulating neuromorphic functionalities with resistive switching memories

Demonstration of Enhanced Switching Variability and Conductance Quantization Properties in a SiO₂ Conducting Bridge Resistive Memory with Embedded Two-Dimensional MoS₂ Material

Artificial Neurons Using Ag−In−Zn−S/Sericin Peptide‐Based Threshold Switching Memristors for Spiking Neural Networks

Contact Info

Product

Resources

About

Low Power Stochastic Neurons From SiO2-Based Bilayer Conductive Bridge Memristors for Probabilistic Spiking Neural Network Applications—Part II: Modeling

Cited by 12 publications

References 34 publications

Material design strategies for emulating neuromorphic functionalities with resistive switching memories

Material design strategies for emulating neuromorphic functionalities with resistive switching memories

Demonstration of Enhanced Switching Variability and Conductance Quantization Properties in a SiO2 Conducting Bridge Resistive Memory with Embedded Two-Dimensional MoS2 Material

Artificial Neurons Using Ag−In−Zn−S/Sericin Peptide‐Based Threshold Switching Memristors for Spiking Neural Networks

Contact Info

Product

Resources

About

Low Power Stochastic Neurons From SiO₂-Based Bilayer Conductive Bridge Memristors for Probabilistic Spiking Neural Network Applications—Part II: Modeling

Demonstration of Enhanced Switching Variability and Conductance Quantization Properties in a SiO₂ Conducting Bridge Resistive Memory with Embedded Two-Dimensional MoS₂ Material