2D-Material-Based Volatile and Nonvolatile Memristive Devices for Neuromorphic Computing

ACS Materials Letters, Journal Year: 2023, Volume and Issue: 5(4), P. 1109 - 1135

Published: March 11, 2023

Neuromorphic computing can process large amounts of information in parallel and provides a powerful tool to solve the von Neumann bottleneck. Constructing an artificial neural network (ANN) is common means realize neuromorphic computing, which has exhibited potential applications pattern recognition, complex sensing, other areas. Reservoir (RC), another approach shown some progress attracted researchers' attention. be generally implemented by fabricating memristive array systems. 2D-material-based systems their ANN RC have been investigated substantially recent years due unique properties these systems, such as atomic-level thickness high carrier mobility. In this Review, we first discuss volatility nonvolatility devices RC. Second, 2D materials that used fabricate are introduced, classification, physical properties, preparation methods presented. Third, working mechanisms synaptic devices, mimicked functions, through Lastly, performance, progress, future development directions analyzed. This work systematically investigates status promotes utilization computing.

Language: Английский

---

Two‐Dimensional Memtransistors for Non‐Von Neumann Computing: Progress and Challenges

Advanced Functional Materials, Journal Year: 2023, Volume and Issue: 34(15)

Published: Oct. 22, 2023

Abstract The increased demand of high‐performance computing systems has exposed the inherent limitations current state‐of‐the‐art von Neumann architecture. Therefore, developing alternate primitives that can offer faster speed with low energy expenditure is critical. In this context, while several non‐volatile memory (NVM) devices such as synaptic transistors, spintronic devices, phase change (PCM), and memristors have been demonstrated in past, their two‐terminal nature necessitates additional peripheral elements increase area overhead. Recently, a new multiterminal device prototype known memtransistor shown tremendous potential to overcome these through exceptional control gate electrostatics enabled by 2D channel materials. perspective, brief overview recent developments 2D‐memtransistor provided, including fundamental operational mechanisms role defects enabling multiple NVM states optical photoresponse. An implementation context neuromorphic, probabilistic, information security, edge‐sensing also provided. Finally, futuristic perspective provided looking toward successful large‐scale technological integration.

Language: Английский

---

Recent Progress in Multiterminal Memristors for Neuromorphic Applications

Advanced Electronic Materials, Journal Year: 2023, Volume and Issue: 9(6)

Published: April 23, 2023

Abstract The essential step for developing neuromorphic systems is to construct more biorealistic elementary devices with rich spatiotemporal dynamics exhibit highly separable responses in dynamic environmental circumstances. Unlike transistor‐based and circuits zeroth‐order complexity, memristors intrinsically express some simple biomimetic functions. However, only two‐terminal structure, precise control of operation principles ensure large space, improved linearity symmetry, multimodal as well high‐order complexity hard achieve a traditional memristor owing its limited degree freedom. Therefore, multiterminal including both concrete terminals virtual (light, pressure, gas, ferroelectric polarity, etc.) have been proposed obtain modulation memristive characteristics. This review focuses on the recent progress their applications. principle, application computing different scenarios, existing challenges future prospects are discussed.

Language: Английский

---

Multi‐Functional Platform for In‐Memory Computing And Sensing Based on 2D Ferroelectric Semiconductor α‐In₂Se₃

Advanced Functional Materials, Journal Year: 2023, Volume and Issue: 34(3)

Published: Oct. 6, 2023

Abstract 2D layered semiconductors with excellent light−matter interaction and atomic‐scale thickness have been envisioned as promising candidates for more than Moore beyond technologies. Here, the first time, a multi‐functional platform is reported that fabricated entirely from wrinkle‐free ferroelectric semiconductor α‐In 2 Se 3 integrated photodetector, reconfigurable logic switching, visual perception processing functions. The intensity‐ wavelength‐dependent resistance used to demodulate broadband optical information into electrical signals perform switching. Moreover, offers dynamically modulated photosensitive sensing in different working modes at pixel level. Using photo‐assisted piezoresponse force microscopy, optical‐engineered polarization switch behavior explored. has sensitivity both stimuli can respond lights visible short‐wavelength infrared region volatile/non‐volatile manner under gate voltage modulation, response of 98 mA W −1 (to 1800 nm light), current on/off ratio over 10 6 , high field‐effect mobility 137.55 cm V s . With its simple structure, unique photoelectric interaction, controllable operating mechanism, potential simplify complexity neuromorphic computing circuitry systems, paving way high‐performance hybrid technologies suitable artificial intelligence applications.

Language: Английский

---

Nanomaterials for Flexible Neuromorphics

Chemical Reviews, Journal Year: 2024, Volume and Issue: 124(22), P. 12738 - 12843

Published: Nov. 5, 2024

The quest to imbue machines with intelligence akin that of humans, through the development adaptable neuromorphic devices and creation artificial neural systems, has long stood as a pivotal goal in both scientific inquiry industrial advancement. Recent advancements flexible electronics primarily rely on nanomaterials polymers owing their inherent uniformity, superior mechanical electrical capabilities, versatile functionalities. However, this field is still its nascent stage, necessitating continuous efforts materials innovation device/system design. Therefore, it imperative conduct an extensive comprehensive analysis summarize current progress. This review highlights applications neuromorphics, involving inorganic (zero-/one-/two-dimensional, heterostructure), carbon-based such carbon nanotubes (CNTs) graphene, polymers. Additionally, comparison summary structural compositions, design strategies, key performance, significant these are provided. Furthermore, challenges future directions pertaining materials/devices/systems associated neuromorphics also addressed. aim shed light rapidly growing attract experts from diverse disciplines (e.g., electronics, science, neurobiology), foster further for accelerated development.

Language: Английский

---

Dopamine detection and integration in neuromorphic devices for applications in artificial intelligence

Device, Journal Year: 2024, Volume and Issue: 2(2), P. 100284 - 100284

Published: Feb. 1, 2024

Language: Английский

---

Roadmap to neuromorphic computing with emerging technologies

APL Materials, Journal Year: 2024, Volume and Issue: 12(10)

Published: Oct. 1, 2024

Language: Английский

---

Probability‐Distribution‐Configurable True Random Number Generators Based on Spin‐Orbit Torque Magnetic Tunnel Junctions

Advanced Science, Journal Year: 2024, Volume and Issue: 11(23)

Published: April 15, 2024

Abstract The incorporation of randomness into stochastic computing can provide ample opportunities for applications such as simulated annealing, non‐polynomial hard problem solving, and Bayesian neuron networks. In these cases, a considerable number random numbers with an accurate configurable probability distribution function (PDF) are indispensable. Preferably, provided at the hardware level to improve speed, efficiency, parallelism. this paper, how spin‐orbit torque magnetic tunnel junctions (SOT‐MTJs) high barriers suitable candidates desired true generators is demonstrated. Not only do SOT‐MTJs perform excellently in speed endurance, but their also be conveniently precisely controlled by writing voltage, which makes them well‐performed Bernoulli bit. By utilizing SOT‐MTJ‐based bits, any PDF, including Gaussian, uniform, exponential, Chi‐square, even arbitrarily defined distributions realized. These PDF‐configurable then promise advance development broaden SOT‐MTJs.

Language: Английский

---

Introducing Chiro-optical Activities in Photonic Synapses for Neuromorphic Computing and In-Memory Logic Operations

ACS Nano, Journal Year: 2024, Volume and Issue: 18(22), P. 14457 - 14468

Published: May 20, 2024

In artificial synaptic devices aimed at mimicking neuromorphic computing systems, electrical or optical pulses, both, are generally used as stimuli. this work, we introduce chiral materials for tailoring the characteristics of photonic to achieve handedness-dependent and in-memory logic gates. based on a pair perovskites, use circularly polarized light (CPL) stimuli mimicked series opto-synaptic functionalities in order emulate multifunctional complex behavior human brain. Upon illumination two-terminal device, anisotropy current has been observed due out-of-plane carrier transport, originating from spin-selective transport. More importantly, gate achieved optoelectronic memristors turned out be chirality-dependent; while an R-device functioned AND gate, device same perovskite opposite chirality (S-device) acted NOR toward operations. These findings perovskite-based synapses can identify further strategies future computing, vision simulation, intelligence.

Language: Английский

---

Bulk photovoltaic effect and high mobility in the polar 2D semiconductor SnP ₂ Se ₆

Science Advances, Journal Year: 2024, Volume and Issue: 10(31)

Published: July 31, 2024

The growth of layered 2D compounds is a key ingredient in finding new phenomena quantum materials, optoelectronics, and energy conversion. Here, we report SnP

Language: Английский

---