Dataset Information

Compact artificial neuron based on anti-ferroelectric transistor.

ABSTRACT: Neuromorphic machines are intriguing for building energy-efficient intelligent systems, where spiking neurons are pivotal components. Recently, memristive neurons with promising bio-plausibility have been developed, but with limited reliability, bulky capacitors or additional reset circuits. Here, we propose an anti-ferroelectric field-effect transistor neuron based on the inherent polarization and depolarization of Hf_0.2Zr_0.8O₂ anti-ferroelectric film to meet these challenges. The intrinsic accumulated polarization/spontaneous depolarization of Hf_0.2Zr_0.8O₂ films implements the integration/leaky behavior of neurons, avoiding external capacitors and reset circuits. Moreover, the anti-ferroelectric neuron exhibits low energy consumption (37 fJ/spike), high endurance (>10¹²), high uniformity and high stability. We further construct a two-layer fully ferroelectric spiking neural networks that combines anti-ferroelectric neurons and ferroelectric synapses, achieving 96.8% recognition accuracy on the Modified National Institute of Standards and Technology dataset. This work opens the way to emulate neurons with anti-ferroelectric materials and provides a promising approach to building high-efficient neuromorphic hardware.

SUBMITTER: Cao R

PROVIDER: S-EPMC9668812 | biostudies-literature | 2022 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Compact artificial neuron based on anti-ferroelectric transistor.

Cao Rongrong R Zhang Xumeng X Liu Sen S Lu Jikai J Wang Yongzhou Y Jiang Hao H Yang Yang Y Sun Yize Y Wei Wei W Wang Jianlu J Xu Hui H Li Qingjiang Q Liu Qi Q

Nature communications 20221117 1

Neuromorphic machines are intriguing for building energy-efficient intelligent systems, where spiking neurons are pivotal components. Recently, memristive neurons with promising bio-plausibility have been developed, but with limited reliability, bulky capacitors or additional reset circuits. Here, we propose an anti-ferroelectric field-effect transistor neuron based on the inherent polarization and depolarization of Hf0.2Zr0.8O2 anti-ferroelectric film to meet th ...[more]

PMID: 36384960

Similar Datasets

Project description:An approximate compact model was developed to provide a convenient method of exploring the initial design space when investigating the performance of micro-electronic devices such as nano-scaled piezoelectronic transistors, where fast ball-park estimates can be very helpful. First of all, the compact model was verified by comparing its predictions with those of accurate axi-symmetric finite element analysis (FEA) using special boundary and interface conditions that enable the replication of the analytical model behaviour. Verification is achieved for a radio frequency (RF) switch and a smaller very-large-scale integrated (VLSI) device, where percentage differences between the compact and FEA model predictions are of the order 10-4 for the RF switch and 10-5 for the VLSI device. This confirms the consistency of complex property data (especially electro-thermo-elastic constants) and geometrical parameter input to both types of models and convincingly demonstrates that the analytical models and FEA for the two devices have been implemented correctly. A second type of boundary and interface condition is also used that is designed to replicate the actual behaviour of the devices in practice. The boundary and interface constraints applied for the verification procedure are relaxed so that there is perfect interface bonding between layers. For this unconstrained case, the resulting deformation is very complex, involving both bending effects and edge effects arising from property mismatches between neighbouring layers. The results for the RF switch show surprisingly good agreement between the predictions of the analytical and FEA results, provided the thickness of the piezoelectric layer is not too thick, implying that the analytical model should help to reduce the parameter design space for such devices. However, for the VLSI device, our results indicate that the compact model leads to much larger errors. For such systems, the compact model is unlikely to be able to reliably reduce the parameter design space, implying that accurate FEA will then need to be used.

Dataset Information

Compact artificial neuron based on anti-ferroelectric transistor.

Publications

Compact artificial neuron based on anti-ferroelectric transistor.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets