A5047981473- Advanced Memory and Neural Computing
- Ferroelectric and Negative Capacitance Devices
- Neural Networks and Reservoir Computing
- Neural dynamics and brain function
- Semiconductor materials and interfaces
- Chalcogenide Semiconductor Thin Films
- CCD and CMOS Imaging Sensors
- Phase-change materials and chalcogenides
- Magnetic properties of thin films
- Neuroscience and Neural Engineering
- Stochastic Gradient Optimization Techniques
- Health Policy Implementation Science
- Quantum and electron transport phenomena
- stochastic dynamics and bifurcation
- Quantum Computing Algorithms and Architecture
- Machine Learning and ELM
- Neural Networks and Applications
- Mental Health and Patient Involvement
- Clinical practice guidelines implementation
- Energy Harvesting in Wireless Networks
- Semiconductor materials and devices
- Topological Materials and Phenomena
- Nonlinear Dynamics and Pattern Formation
Pennsylvania State University
2020-2024
Hong Kong University of Science and Technology
2024
University of Hong Kong
2024
Zunyi Medical University
2022
Uncertainty plays a key role in real-time machine learning. As significant shift from standard deep networks, which does not consider any uncertainty formulation during its training or inference, Bayesian networks are being currently investigated where the network is envisaged as an ensemble of plausible models learnt by Bayes' response to uncertainties sensory data. each synaptic weight sample drawn probability distribution with mean and variance. This letter elaborates on hardware design...
Brain-inspired cognitive computing has so far followed two major approaches - one uses multi-layered artificial neural networks (ANNs) to perform pattern-recognition-related tasks, whereas the other spiking (SNNs) emulate biological neurons in an attempt be as efficient and fault-tolerant brain. While there been considerable progress former area due a combination of effective training algorithms acceleration platforms, latter is still its infancy lack both. SNNs have distinct advantage over...
We report spin-to-charge and charge-to-spin conversion at room temperature in heterostructure devices that interface an archetypal Dirac semimetal, Cd3As2, with a metallic ferromagnet, Ni0.80Fe0.20 (permalloy). The spin-charge interconversion is detected by both spin torque ferromagnetic resonance driven pumping. Analysis of the symmetric anti-symmetric components mixing voltage frequency power dependence pumping signal show behavior these processes consistent previously reported mechanisms...
Probabilistic machine learning enabled by the Bayesian formulation has recently gained significant attention in domain of automated reasoning and decision-making. While impressive strides have been made to scale up performance deep neural networks, they primarily standalone software efforts without any regard underlying hardware implementation. In this article, we propose an "all-spin" network where spintronic provides a better match computing models. To best our knowledge, is first...
Despite the promise of superior efficiency and scalability, real‐world deployment emerging nanoelectronic platforms for brain‐inspired computing have been limited thus far, primarily because inter‐device variations intrinsic non‐idealities. In this work, mitigation these issues is demonstrated by performing learning directly on practical devices through a hardware‐in‐loop approach, utilizing stochastic neurons based heavy metal/ferromagnetic spin–orbit torque heterostructures. The...
Brain-inspired computing—leveraging neuroscientific principles underpinning the unparalleled efficiency of brain in solving cognitive tasks—is emerging to be a promising pathway solve several algorithmic and computational challenges faced by deep learning today. Nonetheless, current research neuromorphic computing is driven our well-developed notions running algorithms on platforms that perform deterministic operations. In this article, we argue taking different route performing temporal...
Emulating various facets of computing principles the brain can potentially lead to development neuro-computers that are able exhibit brain-like cognitive capabilities. In this Letter, we propose a magnetoelectronic neuron utilizes noise as resource and is encode information over time through independent control external voltage signals. We extensively characterize device operation using simulations demonstrate its suitability for neuromorphic platforms performing temporal encoding.
Devices based on the unique phase transitions of change materials (PCMs) like GeTe and Ge2Sb2Te5 (GST) require low-resistance thermally stable Ohmic contacts. This work reviews literature electrical contacts to GeTe, GST, GeCu2Te3 (GCuT), Ge2Cr2Te6 (GCrT), especially due greater number studies. We briefly review how method used measure contact resistance (Rc) specific (ρc) can influence values extracted, since measurements low resistances are susceptible artifacts, we include a direct...
In crossbar array structures, which serves as an "in-memory" compute engine for artificial intelligence (AI) hardware, write sneak path problem causes undesired switching of devices that degrades network accuracy. While custom programming schemes have been proposed, device-level innovations leveraging nonlinear characteristics the cross-point are still under exploration to improve energy eff iciency process. this work, a spintronic device design based on magnetic tunnel junction (MTJ)...
Astrocytes play a central role in inducing concerted phase synchronized neural-wave patterns inside the brain. In this article, we demonstrate that injected radio-frequency signal underlying heavy metal layer of spin-orbit torque oscillator neurons mimic neuron synchronization effect realized by glial cells. Potential application such coupling effects is illustrated context temporal “binding problem.” We also present design coupled neuron-synapse-astrocyte network enabled compact...
Despite the promise of superior efficiency and scalability, real-world deployment emerging nanoelectronic platforms for brain-inspired computing have been limited thus far, primarily because inter-device variations intrinsic non-idealities. In this work, we demonstrate mitigating these issues by performing learning directly on practical devices through a hardware-in-loop approach, utilizing stochastic neurons based heavy metal/ferromagnetic spin-orbit torque heterostructures. We characterize...
First Page
Astrocytes play a central role in inducing concerted phase synchronized neural-wave patterns inside the brain. In this article, we demonstrate that injected radio-frequency signal underlying heavy metal layer of spin-orbit torque oscillator neurons mimic neuron synchronization effect realized by glial cells. Potential application such coupling effects is illustrated context temporal "binding problem". We also present design coupled neuron-synapse-astrocyte network enabled compact...
In crossbar array structures, which serves as an "In-Memory" compute engine for Artificial Intelligence hardware, write sneak path problem causes undesired switching of devices that degrades network accuracy. While custom programming schemes have been proposed, device level innovations leveraging non-linear characteristics the cross-point are still under exploration to improve energy efficiency process. this work, a spintronic design based on Magnetic Tunnel Junction (MTJ) exploiting use...
Brain-inspired computing - leveraging neuroscientific principles underpinning the unparalleled efficiency of brain in solving cognitive tasks is emerging to be a promising pathway solve several algorithmic and computational challenges faced by deep learning today. Nonetheless, current research neuromorphic driven our well-developed notions running algorithms on platforms that perform deterministic operations. In this article, we argue taking different route performing temporal information...
Abstract Background Due to advanced maternal age facing great risks of fertility, such as gestational hypertension, postpartum haemorrhage, miscarriage, etc., we must attach importance care. Guidelines for optimal care are available. It is significance identify potential barriers and tailor practical implementation strategies before implementing the guidelines. The purpose this study aimed use science methods develop that could promote guidelines in clinical practice. Methods Purposive...