A Self-Formed Ag Nanostructure Based Neuromorphic Device Performing Arithmetic Computation and Area Integration: Influence of Presynaptic Pulsing Scheme on Mathematical Precision

ACS Applied Materials & Interfaces, Год журнала: 2025, Номер unknown

Опубликована: Янв. 8, 2025

A material equivalent of a biosynapse is the key to neuromorphic architecture. Here we report self-forming labyrinthine Ag nanostructure activated with few pulses 0.5 V, width and interval set at 50 ms, current compliance (ICC) 400 nA, serving as active for highly stable device programmable volatility. Both conductance (G) its retention time (tr) in potentiated state are found vary linearly pulse number positive negative polarities, nonlinearity factors being noticeably small, ∼0.03 G during potentiation ∼0.08 depression. This was tested over 200 days, results were reproducible. Relying on high linearity, arithmetic operations involving counting integers realized using both often by mixing them feeding sequence. The observed outcomes based independently from tr accurate, deviations typically less than ∼1.5% expected results. Notably, way polarities mixed have an influence, random sequences producing relatively larger integer estimation. However, decreased higher ICC values, which promoted stronger filament formation percolation networks. Besides, values amplitude well, enabled calculation area under curve. Further, exhibited simulation-based image classification accuracy 94.95%, close ideal value (96.05%). Simulations utilizing finite element method showcased uniqueness morphology, giving rise field intensification along potential percolative paths.

Язык: Английский

---

Overview of emerging electronics technologies for artificial intelligence: A review

Materials Today Electronics, Год журнала: 2025, Номер unknown, С. 100136 - 100136

Опубликована: Янв. 1, 2025

Язык: Английский

---

Flash Memory for Synaptic Plasticity in Neuromorphic Computing: A Review

Biomimetics, Год журнала: 2025, Номер 10(2), С. 121 - 121

Опубликована: Фев. 18, 2025

The rapid expansion of data has made global access easier, but it also demands increasing amounts energy for storage and processing. In response, neuromorphic electronics, inspired by the functionality biological neurons synapses, have emerged as a growing area research. These devices enable in-memory computing, helping to overcome "von Neumann bottleneck", limitation caused separation memory processing units in traditional von architecture. By leveraging multi-bit non-volatility, biologically features, Ohm's law, synaptic show great potential reducing consumption multiplication accumulation operations. Within various non-volatile technologies available, flash stands out highly competitive option storing large volumes data. This review highlights recent advancements computing that utilize NOR, AND, NAND memory. delves into array architecture, operational methods, electrical properties memory, emphasizing its application different neural network designs. providing detailed overview memory-based this offers valuable insights optimizing use across diverse applications.

Язык: Английский

---

Microstructure-modulated conductive filaments in Ruddlesden-Popper perovskite-based memristors and their application in artificial synapses

Materials Today Physics, Год журнала: 2025, Номер unknown, С. 101708 - 101708

Опубликована: Март 1, 2025

Язык: Английский

---

Smart Dust for Chemical Mapping

Advanced Materials, Год журнала: 2025, Номер unknown

Опубликована: Март 25, 2025

This review article explores the transformative potential of smart dust systems by examining how existing chemical sensing technologies can be adapted and advanced to realize their full capabilities. Smart dust, characterized submillimeter-scale autonomous platforms, offers unparalleled opportunities for real-time, spatiotemporal mapping across diverse environments. introduces technological advancements underpinning these systems, critically evaluates current limitations, outlines new avenues development. Key challenges, including multi-compound detection, system control, environmental impact, cost, are discussed alongside solutions. By leveraging innovations in miniaturization, wireless communication, AI-driven data analysis, sustainable materials, this highlights promise address critical challenges monitoring, healthcare, agriculture, defense sectors. Through lens, provides a strategic roadmap advancing from concept practical application, emphasizing its role transforming understanding management complex systems.

Язык: Английский

---

Metal oxide photoelectric synaptic transistor with CeOx floating gate and its application in neuromorphic computing

Applied Physics Letters, Год журнала: 2025, Номер 126(19)

Опубликована: Май 12, 2025

Photoelectric synaptic transistors (PSTs) based on metal oxide semiconductors (MOSs) have shown promising applications in visual perception and photonic computing. However, the response range of PST is limited ultra-violet region due to wide bandgap MOS. Herein, a visible light-driven InGaZnO CeOx floating gate presented. The optical improved introduction oxygen vacancies gate, tunable characteristics are endowed. Various behaviors under light stimulation been simulated, including paired-pulse facilitation, high-pass filtering characteristics, transition from short-term memory long-term memory, learning-experience behavior. multilevel conductance modulation realized through programming electrical erasing operations. An artificial neural network was constructed plasticity PST, 95.3% accuracy achieved image recognition. This work promotes development oxide-based provides candidate for bionics inspired by light.

Язык: Английский

---

Ion-mediated oxide transistors for neuromorphic electronics: Materials, devices, and perspectives

Applied Physics Reviews, Год журнала: 2025, Номер 12(2)

Опубликована: Май 16, 2025

Constrained by the physical architecture of von Neumann computing with separated storage and computation, neuromorphic architectures have been proposed. Ion-mediated oxide synaptic transistors (IOSTs), their unique biomimetic characteristics, become a key fundamental component in construction systems. This review comprehensively explores principles critical indicators IOSTs discusses advantages recent developments associated employing various electrolyte materials as dielectric layers IOSTs, including ionic liquids, gels, organic polymer electrolytes, inorganic solid electrolytes. Furthermore, we explore extensive applications across multiple domains, such multisensory bionics computing. article provides an exhaustive perspective on research related to system integration applications, offering insights into evolving landscape electronics.

Язык: Английский

---

Wavelength-selective synaptic devices based on Graphdiyne/WSe2 for multi-color image recognition and neuromorphic computing

Chemical Engineering Journal, Год журнала: 2025, Номер unknown, С. 164215 - 164215

Опубликована: Июнь 1, 2025

Язык: Английский

---