A5013365606- Advanced Memory and Neural Computing
- Ferroelectric and Negative Capacitance Devices
- Magnetic properties of thin films
- Neural Networks and Reservoir Computing
- Neural Networks and Applications
- 2D Materials and Applications
- Semiconductor materials and devices
- Magnetic Properties and Applications
- Quantum and electron transport phenomena
- Advancements in Semiconductor Devices and Circuit Design
- Graphene research and applications
- Cellular Automata and Applications
- Quantum-Dot Cellular Automata
- ZnO doping and properties
- Modular Robots and Swarm Intelligence
- MXene and MAX Phase Materials
- Advanced Data Storage Technologies
- Graphene and Nanomaterials Applications
- Perovskite Materials and Applications
- Photoreceptor and optogenetics research
- Metallic Glasses and Amorphous Alloys
- Neuroscience and Neural Engineering
- Theoretical and Computational Physics
- Quantum Computing Algorithms and Architecture
- Electrochemical sensors and biosensors
The University of Texas at Austin
2017-2025
University of California, Berkeley
2024
The University of Texas at Dallas
2021-2022
Weatherford College
2022
Arizona State University
2022
Sandia National Laboratories
2021-2022
Applied Materials (Germany)
2022
Sandia National Laboratories California
2022
Applied Materials (United States)
2021
Stanford University
2016-2019
CMOS-based computing systems that employ the von Neumann architecture are relatively limited when it comes to parallel data storage and processing. In contrast, human brain is a living computational signal processing unit operates with extreme parallelism energy efficiency. Although numerous neuromorphic electronic devices have emerged in last decade, most of them rigid or contain materials toxic biological systems. this work, we report on biocompatible bilayer graphene-based artificial...
Abstract In the ‘Beyond Moore’s Law’ era, with increasing edge intelligence, domain-specific computing embracing unconventional approaches will become increasingly prevalent. At same time, adopting a variety of nanotechnologies offer benefits in energy cost, computational speed, reduced footprint, cyber resilience, and processing power. The time is ripe for roadmap to guide future research, this collection aims fill that need. authors provide comprehensive neuromorphic using electron spins,...
Abstract Spintronic computing promises superior energy efficiency and nonvolatility compared to conventional field-effect transistor logic. But, it has proven difficult realize spintronic circuits with a versatile, scalable device design that is adaptable emerging material physics. Here we present prototypes of logic encode information in the position magnetic domain wall ferromagnetic wire. We show single three-terminal can perform inverter buffer operations. demonstrate one drive two...
The brain has effectively proven a powerful inspiration for the development of computing architectures in which processing is tightly integrated with memory, communication event-driven, and analog computation can be performed at scale. These neuromorphic systems increasingly show an ability to improve efficiency speed scientific artificial intelligence applications. Herein, it proposed that brain's ubiquitous stochasticity represents additional source expanding reach probabilistic To date,...
Abstract In neuromorphic computing, artificial synapses provide a multi‐weight (MW) conductance state that is set based on inputs from neurons, analogous to the brain. Herein, magnetic materials use tunnel junction (MTJ) and domain wall (DW) are explored. By fabricating lithographic notches in DW track underneath single MTJ, 3–5 stable resistance states can be repeatably controlled electrically using spin‐orbit torque achieved. The effect of geometry synapse behavior explored, showing...
The development of an efficient neuromorphic computing system requires the use nanodevices that intrinsically emulate biological behavior neurons and synapses. While numerous artificial synapses have been shown to store weights in a manner analogous synapses, challenge developing neuron is impeded by necessity include leaking, integrating, firing, lateral inhibition features. In particular, previous proposals for required external circuits perform inhibition, thereby decreasing efficiency...
Inspired by the parallelism and efficiency of brain, several candidates for artificial synapse devices have been developed neuromorphic computing, yet a nonlinear asymmetric synaptic response curve precludes their use backpropagation, foundation modern supervised learning. Spintronic devices—which benefit from high endurance, low power consumption, latency, CMOS compatibility—are promising technology memory, domain-wall magnetic tunnel junction (DW-MTJ) shown to implement functions such as...
The spatiotemporal nature of neuronal behavior in spiking neural networks (SNNs) make SNNs promising for edge applications that require high energy efficiency. To realize hardware, spintronic neuron implementations can bring advantages scalability and Domain wall (DW) based magnetic tunnel junction (MTJ) devices are well suited probabilistic given their intrinsic integrate-and-fire with tunable stochasticity. Here, we present a scaled DW-MTJ voltage-dependent firing probability. measured was...
Spiking neural networks seek to emulate biological computation through interconnected artificial neuron and synapse devices. Spintronic neurons can leverage magnetization physics mimic functions, such as integration tied magnetic domain wall (DW) propagation in a patterned nanotrack firing the resistance change of tunnel junction (MTJ), captured wall-magnetic (DW-MTJ) device. Leaking, relaxation when it is not under stimulation, also predicted be implemented based on DW drift relaxes low...
Ambipolar dual-gate transistors based on low-dimensional materials, such as graphene, carbon nanotubes, black phosphorus, and certain transition metal dichalcogenides (TMDs), enable reconfigurable logic circuits with a suppressed off-state current. These achieve the same logical output complementary metal–oxide semiconductor (CMOS) fewer offer greater flexibility in design. The primary challenge lies cascadability power consumption of these gates static CMOS-like connections. In this...
Emerging devices, such as magnetic tunnel junctions, are key for energy-efficient, performant future computing systems. However, designing devices with the desirable specification and performance these applications is often found to be time-consuming non-trivial. Here, we investigate design optimization of spin–orbit torque spin transfer junction models probabilistic true random number generation. We leverage reinforcement learning evolutionary vary device material properties various...
On‐chip integration of inductors and transformers can enable power converters with high frequency, control bandwidth, low interconnect loss for high‐current computing applications. Nevertheless, depositing high‐quality magnetic materials that are back end line compatible complementary metal‐oxide‐semiconductor (CMOS) technology, affordable (fast deposition), but optimized these applications is challenging. Nanogranular materials, such as CoZrO, promising candidates core on‐chip components...
Black phosphorus (BP) is a promising two-dimensional (2D) material for nanoscale transistors, due to its expected higher mobility than other 2D semiconductors. While most studies have reported ambipolar BP with stronger p-type transport, it important fabricate both unipolar p- and n-type transistors low-power digital circuits. Here, we report low work function Sc Er contacts, demonstrating record high current of 200 μA/μm in 6.5 nm thick BP. Intriguingly, the electrical transport...
Spintronic three-terminal magnetic-tunnel-junction (3T-MTJ) devices have gained considerable interest in the field of neuromorphic computing. Previously, these required external circuitry to implement leaking functionality that leaky integrate-and-fire (LIF) neurons should display. However, use results decreased device efficiency. We previously demonstrated lateral inhibition with a 3T-MTJ neuron intrinsically performs leaking, integrating, and firing functions; however, it fabrication...
Abstract Topological solitons are exciting candidates for the physical implementation of next-generation computing systems. As these nanoscale and can be controlled with minimal energy consumption, they ideal to fulfill emerging needs in era big data processing storage. Magnetic domain walls (DWs) magnetic skyrmions two types topological that particularly systems light their non-volatility, scalability, rich interactions, ability exhibit non-linear behaviors. Here we summarize development...
We report a combined directing effect of the simultaneously applied graphoepitaxy and electric field on self-assembly cylinder forming polystyrene-b-poly(dimethylsiloxane) block copolymer in thin films. A correlation length up to 20 μm uniaxial ordered striped patterns is an order magnitude greater than that produced by either or alignment alone achieved at reduced annealing times. The angle between direction topographic guides as well dimensions trenches affected both quality ordering...
There have been recent efforts towards the development of biologically-inspired neuromorphic devices and architecture. Here, we show a synapse circuit that is designed to perform spike-timing-dependent plasticity which works with leaky, integrate, fire neuron in computing The consists three-terminal magnetic tunnel junction mobile domain wall between two low-pass filters has modeled SPICE. results current flowing through highly correlated timing delay pre-synaptic post-synaptic neurons....
We investigate the valley Hall effect (VHE) in monolayer WSe2 field-effect transistors using optical Kerr rotation measurements at 20 K. While studies of VHE have so far focused on n-doped MoS2, we observe both n- and p-doping regimes. Hole doping enables access to large spin-splitting valence band this material. The probe spatial distribution carrier imbalance induced by VHE. Under current flow, distinct spin-valley polarization along edges transistor channel. From analysis magnitude...
Bayesian neural networks (BNNs) combine the generalizability of deep (DNNs) with a rigorous quantification predictive uncertainty, which mitigates overfitting and makes them valuable for high-reliability or safety-critical applications. However, probabilistic nature BNNs more computationally intensive on digital hardware so far, less directly amenable to acceleration by analog in-memory computing as compared DNNs. This work exploits novel spintronic bit cell that efficiently compactly...
Complementary metal oxide semiconductor (CMOS) devices display volatile characteristics, and are not well suited for analog applications such as neuromorphic computing. Spintronic devices, on the other hand, exhibit both non-volatile features, which well-suited to Consequently, these novel at forefront of beyond-CMOS artificial intelligence applications. However, a large quantity still require use CMOS, decreases efficiency system. To resolve this, we have previously proposed number neurons...
Abstract Perpendicular magnetic tunnel junction (pMTJ)-based true-random number generators (RNGs) can consume orders of magnitude less energy per bit than CMOS pseudo-RNGs. Here, we numerically investigate with a macrospin Landau–Lifshitz-Gilbert equation solver the use pMTJs driven by spin–orbit torque to directly sample numbers from arbitrary probability distributions help tunable tree. The tree operates dynamically biasing sequences pMTJ relaxation events, called ‘coinflips’, via an...