Search Results (5)

Neuromorphic computing inspired by biological synapses requires memory devices capable of mimicking short-term memory (STM) and associative learning. In this study, we investigate a 15 nm-thick Hafnium zirconium oxide (HZO)-based ferroelectric memristor device, which exhibits robust STM characteristics and successfully replicates Pavlov’s dog experiment. The optimized 15 nm HZO layer demonstrates enhanced ferroelectric properties, including a stable orthorhombic phase and a reliable short-term synaptic response. Furthermore, through a series of conditional learning experiments, the device effectively reproduces associative learning by forming and extinguishing conditioned responses, closely resembling biological neural plasticity. The number of training repetitions significantly affects the retention of learned responses, indicating a transition from STM-like behavior to longer-lasting memory effects. These findings highlight the potential of the optimized ferroelectric device in neuromorphic applications, particularly for implementing real-time learning and memory in artificial intelligence systems. Full article

Biological systems learn from past experiences by establishing relationships between two simultaneously occurring events, a phenomenon known as associative learning. This concept has promising applications in modern AI (Artificial Intelligence) and ML (Machine Learning). To leverage it effectively, a precise electrical model that can simulate associative learning observed in biological systems is essential. The paper focuses on modeling Pavlov’s famous experiment related to the drooling of dogs at the sound of bell after associating the food with the bell during training. The study addresses limitations in existing circuit designs that fail to accurately replicate associative learning in dogs, particularly when the sequence of food and bell signals deviates from a specific pattern. We propose a novel design using a few CMOS (Complementary Metal Oxide Semiconductor) transistors and memristor models that produces an output corresponding to the dogs drooling only when food and bell signals are associated, mirroring real-life training conditions. The results section first discusses simulations using the standard TiO₂ (Titanium Oxide) memristor model, followed by experimental results obtained from a classical memristor emulator. Both simulation and experimental findings confirm the effectiveness of the circuit designs. Full article

The development of brain-inspired spiking neural networks (SNNs) has great potential for neuromorphic edge computing applications, while challenges remain in optimizing power-efficiency and silicon utilization. Neurons, synapses and spike-based learning algorithms form the fundamental information processing mechanism of SNNs. In an effort to achieve compact and biologically plausible SNNs while restricting power consumption, we propose a set of new neuromorphic building circuits, including an analog Leaky Integrate-and-Fire (LIF) neuron circuit, configurable synapse circuits and Spike Driven Synaptic Plasticity (SDSP) learning algorithm circuits. Specifically, we explore methods to minimize large leakage current and device mismatch effects, and optimize the design of these neuromorphic circuits to enable low-power operation. A reconfigurable mixed-signal SNN is proposed based on the building circuits, allowing flexible configuration of synapse weights and attributes, resulting in enhanced SNN functionality and reduced unnecessary power consumption. This SNN chip is fabricated using 55 nm CMOS technology, and test results indicate that the proposed circuits have the ability to closely mimic the behaviors of LIF neurons, synapses and SDSP mechanisms. By configuring synaptic arrays, we established varied connections between neurons in the SNN and demonstrated that this SNN chip can implement Pavlov’s dog associative learning and binary classification tasks, while dissipating less energy per spike of the order of Pico Joules per spike at a firing rate ranging from 30 Hz to 1 kHz. The proposed circuits can be used as building blocks for constructing large-scale SNNs in neuromorphic processors. Full article

This review article attempts to provide a comprehensive review of the recent progress in cerium oxide (CeO₂)-based resistive random-access memories (RRAMs). CeO₂ is considered the most promising candidate because of its multiple oxidation states (Ce³⁺ and Ce⁴⁺), remarkable resistive-switching (RS) uniformity in DC mode, gradual resistance transition, cycling endurance, long data-retention period, and utilization of the RS mechanism as a dielectric layer, thereby exhibiting potential for neuromorphic computing. In this context, a detailed study of the filamentary mechanisms and their types is required. Accordingly, extensive studies on unipolar, bipolar, and threshold memristive behaviors are reviewed in this work. Furthermore, electrode-based (both symmetric and asymmetric) engineering is focused for the memristor’s structures such as single-layer, bilayer (as an oxygen barrier layer), and doped switching-layer-based memristors have been proved to be unique CeO₂-based synaptic devices. Hence, neuromorphic applications comprising spike-based learning processes, potentiation and depression characteristics, potentiation motion and synaptic weight decay process, short-term plasticity, and long-term plasticity are intensively studied. More recently, because learning based on Pavlov’s dog experiment has been adopted as an advanced synoptic study, it is one of the primary topics of this review. Finally, CeO₂-based memristors are considered promising compared to previously reported memristors for advanced synaptic study in the future, particularly by utilizing high-dielectric-constant oxide memristors. Full article

The development of synthetic biology has enabled massive progress in biotechnology and in approaching research questions from a brand-new perspective. In particular, the design and study of gene regulatory networks in vitro, in vivo, and in silico have played an increasingly indispensable role in understanding and controlling biological phenomena. Among them, it is of great interest to understand how associative learning is formed at the molecular circuit level. Mathematical models are increasingly used to predict the behaviours of molecular circuits. Fernando’s model, which is one of the first works in this line of research using the Hill equation, attempted to design a synthetic circuit that mimics Hebbian learning in a neural network architecture. In this article, we carry out indepth computational analysis of the model and demonstrate that the reinforcement effect can be achieved by choosing the proper parameter values. We also construct a novel circuit that can demonstrate forced dissociation, which was not observed in Fernando’s model. Our work can be readily used as reference for synthetic biologists who consider implementing circuits of this kind in biological systems. Full article