Neuro-Inspired-in-Memory Computing Using Charge-Trapping MemTransistor on Germanium as Synaptic Device

Chou, Yi-Teh; Tsai, Chien-Wei; Yi, Chin-Ya; Chung, Wan-Hsuan; Wang, Shin-Yuan; Chien, Chao–Hsin

doi:10.1109/ted.2020.3008887

Cited by 8 publications

(7 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…8 Compared to (ref. 48), we achieved 85% accuracy with as low as 20 conductance states providing a reasonable balance between accuracy and variation margin (DG $9). It should be noted that it is unclear what NL and DG values are appropriate for the reliable operation of a neuromorphic system.…”

Section: (C) Conductance Weight States (N States )mentioning

confidence: 87%

An artificial synaptic transistor using an α-In₂Se₃ van der Waals ferroelectric channel for pattern recognition

et al. 2021

View full text Add to dashboard Cite

show abstract

Section: (C) Conductance Weight States (N States )mentioning

confidence: 87%

An artificial synaptic transistor using an α-In₂Se₃ van der Waals ferroelectric channel for pattern recognition

et al. 2021

View full text Add to dashboard Cite

show abstract

“…4 (a) and (b). Comparing with the case of HfO2 as reference [31], using HfTiO as CTL can enlarge the memory windows. In the case of H9T1, H19T1, H29T1 and HT, the memory window is enlarged by 430.7, 504.8, 488.7 and 593.5 mV.…”

Section: Physical and Memory Characteristics Of Charge Trapping mentioning

confidence: 99%

“…Even though these candidates have been shown obvious advantage in the ANN applications, however, the degradation of recognition accuracy caused by the nonlinearity of weight-to-pulse and [26,28,29]. Therefore, we have proposed the Charge Trapping MemTransistor (CTMT) beforehand [30,31] which has comprehensive improvements in the abovementioned criteria.…”

Section: Introductionmentioning

confidence: 99%

“…We then fabricated CTMTs using Ti-doped HfO2 as CTL and characterized their synaptic characteristics such as the nonlinearity of pulse-number-to-weight relation, number of weight states and conductance on/off ratio. Finally, we utilized the NeuroSim [38,39] simulator with MLP and CNN as ANN models to benchmark the CTMT as synaptic device against other types of synaptic device [19][20][21][22][23]26] and our previous CTMT [31].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Improving the Performance of Charge Trapping Memtransistor as Synaptic Device by Ti-Doped HfO₂

Chou

Chung

Tsai

et al. 2021

IEEE J. Electron Devices Soc.

Self Cite

View full text Add to dashboard Cite

“…Memristor is the fourth basic electronic component after resistance, capacitance, and inductance. [ 1–4 ] It has the memory characteristic of continuous regulation of resistance, which is highly similar to the synaptic plasticity function of brain cognition. Thus, the memristor has been considered as an important hardware basis for constructing advanced neuromorphic computing and is further expected to fundamentally break through the bottleneck of traditional von Neumann computer system architecture and greatly improve data processing speed and energy efficiency.…”

Section: Introductionmentioning

confidence: 99%

Modulate the Interfacial Oxygen Vacancy for High Performance MoS₂ Photosensing−Memory Device

Peng

Luo

Liang

et al. 2022

Advanced Optical Materials

View full text Add to dashboard Cite

Developing memristor with photosensing behavior would facilitate the construction of an advanced neuromorphic computing framework for future sensing−memory−computing integrated system. Although the memristor has already got much attention, its optoelectronic sensing mechanism is still unclear, thus the enhancement is challenging. Herein, a MoS2 memristor is fabricated on the oxygen‐vacancy‐rich Bi1.5MgNb1.5O7 thin film. By combining the voltage stress and photoexcitation, the memristor, its photosensing property, and their correlation have been systematically studied. The results show that the interfacial oxygen vacancy can trap the photogenerated electron obeying the photogating effect which tunes the Fermi level of MoS2 and exhibits tunable hysteresis behavior. The photoexcitation mechanism has been studied showing wavelength and voltage‐dependent photosensing−memory performance. The responsivity can reach up to 2 × 1011 V W−1 at 550 nm illumination and the hysteresis ratio (ΔH/Vg), one of the factors determining the memory, can be modulated by 40 times, which is larger than for previous 2D materials‐based memristors. Contour maps of external quantum efficiency show illumination, voltage, and wavelength dependence where the value becomes larger at lower illumination power density and higher voltage. This study reveals the interfacial enhanced photosensing mechanism of MoS2 device and provides new way to enrich memristor performance.

show abstract

Neuro-Inspired-in-Memory Computing Using Charge-Trapping MemTransistor on Germanium as Synaptic Device

Cited by 8 publications

References 20 publications

An artificial synaptic transistor using an α-In₂Se₃ van der Waals ferroelectric channel for pattern recognition

An artificial synaptic transistor using an α-In₂Se₃ van der Waals ferroelectric channel for pattern recognition

Improving the Performance of Charge Trapping Memtransistor as Synaptic Device by Ti-Doped HfO₂

Modulate the Interfacial Oxygen Vacancy for High Performance MoS₂ Photosensing−Memory Device

Contact Info

Product

Resources

About

Neuro-Inspired-in-Memory Computing Using Charge-Trapping MemTransistor on Germanium as Synaptic Device

Cited by 8 publications

References 20 publications

An artificial synaptic transistor using an α-In2Se3 van der Waals ferroelectric channel for pattern recognition

An artificial synaptic transistor using an α-In2Se3 van der Waals ferroelectric channel for pattern recognition

Improving the Performance of Charge Trapping Memtransistor as Synaptic Device by Ti-Doped HfO2

Modulate the Interfacial Oxygen Vacancy for High Performance MoS2 Photosensing−Memory Device

Contact Info

Product

Resources

About

An artificial synaptic transistor using an α-In₂Se₃ van der Waals ferroelectric channel for pattern recognition

An artificial synaptic transistor using an α-In₂Se₃ van der Waals ferroelectric channel for pattern recognition

Improving the Performance of Charge Trapping Memtransistor as Synaptic Device by Ti-Doped HfO₂

Modulate the Interfacial Oxygen Vacancy for High Performance MoS₂ Photosensing−Memory Device