Still searching Arxiv, refresh for possibly better results.

total 4960took 0.11s

Chaos-guided Input Structuring for Improved Learning in Recurrent Neural NetworksDec 26 2017Feb 18 2018Anatomical studies demonstrate that brain reformats input information to generate reliable responses for performing computations. However, it remains unclear how neural circuits encode complex spatio-temporal patterns. We show that neural dynamics are ... More

Encoding Neural and Synaptic Functionalities in Electron Spin: A Pathway to Efficient Neuromorphic ComputingNov 07 2017Dec 20 2017Present day computers expend orders of magnitude more computational resources to perform various cognitive and perception related tasks that humans routinely perform everyday. This has recently resulted in a seismic shift in the field of computation where ... More

Image Edge Detection based on Swarm Intelligence using Memristive NetworksJun 16 2016Feb 14 2017Recent advancements in the development of memristive devices has opened new opportunities for hardware implementation of non-Boolean computing. To this end, the suitability of memristive devices for swarm intelligence algorithms has enabled researchers ... More

Ising spin model using Spin-Hall Effect (SHE) induced magnetization reversal in Magnetic-Tunnel-JunctionSep 19 2016Sep 25 2016Ising spin model is considered as an efficient computing method to solve combinatorial optimization problems based on its natural tendency of convergence towards low energy state. The underlying basic functions facilitating the Ising model can be categorized ... More

Capacitively Driven Global Interconnect with Magnetoelectric Switching Based Receiver for Higher Energy EfficiencyFeb 26 2018We propose capacitively driven low-swing global interconnect circuit using a receiver that utilizes magnetoelectric (ME) effect induced magnetization switching to reduce the energy consumption. Capacitively driven wire has recently been shown to be effective ... More

Stochastic Spiking Neural Networks Enabled by Magnetic Tunnel Junctions: From Nontelegraphic to Telegraphic Switching RegimesSep 26 2017Jan 26 2018Stochastic spiking neural networks based on nanoelectronic spin devices can be a possible pathway to achieving "brainlike" compact and energy-effcient cognitive intelligence. The computational model attempt to exploit the intrinsic device stochasticity ... More

Magnetic Tunnel Junction Mimics Stochastic Cortical Spiking NeuronsOct 01 2015Dec 19 2017Brain-inspired computing architectures attempt to mimic the computations performed in the neurons and the synapses in the human brain in order to achieve its efficiency in learning and cognitive tasks. In this work, we demonstrate the mapping of the probabilistic ... More

Hierarchical Temporal Memory Based on Spin-Neurons and Resistive Memory for Energy-Efficient Brain-Inspired ComputingFeb 10 2014Hierarchical temporal memory (HTM) tries to mimic the computing in cerebral-neocortex. It identifies spatial and temporal patterns in the input for making inferences. This may require large number of computationally expensive tasks like, dot-product evaluations. ... More

Significance Driven Hybrid 8T-6T SRAM for Energy-Efficient Synaptic Storage in Artificial Neural NetworksFeb 27 2016Multilayered artificial neural networks (ANN) have found widespread utility in classification and recognition applications. The scale and complexity of such networks together with the inadequacies of general purpose computing platforms have led to a significant ... More

Image Edge Detection based on Swarm Intelligence using Memristive NetworksJun 16 2016Recent advancements in the development of memristive devices has opened new opportunities for hardware implementation of non-Boolean computing. To this end, the suitability of memristive devices for swarm intelligence algorithms has enabled researchers ... More

Attention Tree: Learning Hierarchies of Visual Features for Large-Scale Image RecognitionAug 01 2016One of the key challenges in machine learning is to design a computationally efficient multi-class classifier while maintaining the output accuracy and performance. In this paper, we present a tree-based classifier: Attention Tree (ATree) for large-scale ... More

MESL: Proposal for a Non-volatile Cascadable Magneto-Electric Spin LogicNov 23 2016In the quest for novel, scalable and energy-efficient computing technologies, many non-charge based logic devices are being explored. Recent advances in multi-ferroic materials have paved the way for electric field induced low energy and fast switching ... More

Implicit Generative Modeling of Random Noise during Training for Adversarial RobustnessJul 05 2018Feb 08 2019We introduce a Noise-based prior Learning (NoL) approach for training neural networks that are intrinsically robust to adversarial attacks. We find that the implicit generative modeling of random noise with the same loss function used during posterior ... More

Spintronic Switches for Ultra Low Energy On-Chip and Inter-Chip Current-Mode InterconnectsApr 08 2013Apr 19 2013Energy-efficiency and design-complexity of high-speed on-chip and inter-chip data-interconnects has emerged as the major bottleneck for high-performance computing-systems. As a solution, we propose an ultra-low energy interconnect design-scheme using ... More

ReStoCNet: Residual Stochastic Binary Convolutional Spiking Neural Network for Memory-Efficient Neuromorphic ComputingFeb 11 2019In this work, we propose ReStoCNet, a residual stochastic multilayer convolutional Spiking Neural Network (SNN) composed of binary kernels, to reduce the synaptic memory footprint and enhance the computational efficiency of SNNs for complex pattern recognition ... More

Unsupervised Regenerative Learning of Hierarchical Features in Spiking Deep Networks for Object RecognitionFeb 03 2016We present a spike-based unsupervised regenerative learning scheme to train Spiking Deep Networks (SpikeCNN) for object recognition problems using biologically realistic leaky integrate-and-fire neurons. The training methodology is based on the Auto-Encoder ... More

Short-Term Plasticity and Long-Term Potentiation in Magnetic Tunnel Junctions: Towards Volatile SynapsesOct 31 2015Feb 01 2016Synaptic memory is considered to be the main element responsible for learning and cognition in humans. Although traditionally non-volatile long-term plasticity changes have been implemented in nanoelectronic synapses for neuromorphic applications, recent ... More

Inclusive prompt photon production in electron-nucleus scattering at small xFeb 26 2018We compute the differential cross-section for inclusive prompt photon production in deeply inelastic scattering (DIS) of electrons on nuclei at small $x$ in the framework of the Color Glass Condensate (CGC) effective theory. The leading order (LO) computation ... More

Short-Term Plasticity and Long-Term Potentiation in Magnetic Tunnel Junctions: Towards Volatile SynapsesOct 31 2015Dec 19 2017Synaptic memory is considered to be the main element responsible for learning and cognition in humans. Although traditionally non-volatile long-term plasticity changes have been implemented in nanoelectronic synapses for neuromorphic applications, recent ... More

Exploring Ultra Low-Power on-Chip Clocking Using Functionality Enhanced Spin-Torque SwitchesDec 30 2013Emerging spin-torque (ST) phenomena may lead to ultra-low-voltage, high-speed nano-magnetic switches. Such current-based-switches can be attractive for designing low swing global-interconnects, like, clocking-networks and databuses. In this work we present ... More

Quantum theory of light emission from quantum dots coupled to structured photonic reservoirs and acoustic phononsApr 13 2015Electron-phonon coupling in semiconductor quantum dots plays a significant role in determining the optical properties of excited excitons, especially the spectral nature of emitted photons. This paper presents a comprehensive theory and analysis of emission ... More

Magnetic Skyrmions for Cache MemoryMay 01 2017Magnetic skyrmions (MS) are particle-like spin structures with whirling configuration, which are promising candidates for spin-based memory. MS contains alluring features including remarkably high stability, ultra low driving current density, and compact ... More

Theory of phonon-modified quantum dot photoluminescence intensity in structured photonic reservoirsNov 21 2014Apr 10 2015The spontaneous emission rate of a quantum dot coupled to a structured photonic reservoir is determined by the frequency dependence of its local density of photon states. Through phonon-dressing, a breakdown of Fermi's golden rule can occur for certain ... More

Spontaneous emission from a quantum dot in a structured photonic reservoir: phonon-mediated breakdown of Fermi's golden ruleJun 13 2014Jun 01 2015We describe how a structured photonic medium controls the spontaneous emission rate from an excited quantum dot in the presence of electron-phonon coupling. We analyze this problem using a polaron transformed master equation and we consider specific examples ... More

Stimulating STDP to Exploit Locality for Lifelong Learning without Catastrophic ForgettingFeb 08 2019Stochastic gradient descent requires that training samples be drawn from a uniformly random distribution of the data. For a deployed system that must learn online from an uncontrolled and unknown environment, the ordering of input samples often fails ... More

Investigation of Dependence between Time-zero and Time-dependent Variability in High-k NMOS TransistorsAug 17 2016Bias Temperature Instability (BTI) is a major reliability concern in CMOS technology, especially with High dielectric constant (High-\k{appa}/HK) metal gate (MG) transistors. In addition, the time independent process induced variation has also increased ... More

Tree-CNN: A Hierarchical Deep Convolutional Neural Network for Incremental LearningFeb 15 2018May 23 2018In recent years, Convolutional Neural Networks (CNNs) have shown remarkable performance in many computer vision tasks such as object recognition and detection. However, complex training issues, such as `catastrophic forgetting' and hyper-parameter tuning, ... More

Technology Aware Training in Memristive Neuromorphic Systems based on non-ideal Synaptic CrossbarsNov 24 2017The advances in the field of machine learning using neuromorphic systems have paved the pathway for extensive research on possibilities of hardware implementations of neural networks. Various memristive technologies such as oxide-based devices, spintronics ... More

Uniqueness of positive solution for a quasilinear elliptic equation in heisenberg groupDec 09 2015In this article we are interested in addressing the question of existence and uniqueness of positive solution to a quasilinear elliptic equation involving p-laplacian in Heisenberg Group. The idea is to prove the uniqueness by using Diaz-Saa Inequality ... More

Behavioral response to strong aversive stimuli: A neurodynamical modelApr 04 2007In this paper a theoretical model of functioning of a neural circuit during a behavioral response has been proposed. A neural circuit can be thought of as a directed multigraph whose each vertex is a neuron and each edge is a synapse. It has been assumed ... More

An FFT based measure of phase synchronizationDec 04 2006Apr 25 2008In this paper phase of a signal has been viewed from a different angle. According to this view a signal can have countably infinitely many phases, one associated with each Fourier component. In other words each frequency has a phase associated with it. ... More

A structural and a functional aspect of stable information processing by the brainJan 21 2007Jun 16 2007In this paper a model of neural circuit in the brain has been proposed which is composed of cyclic sub-circuits. A big loop has been defined to be consisting of a feed forward path from the sensory neurons to the highest processing area of the brain and ... More

A new measure of phase synchronization for a pair of time series and seizure focus localizationDec 13 2006Dec 22 2006Defining and measuring phase synchronization in a pair of nonlinear time series are highly nontrivial. This can be done with the help of Fourier transform, when it exists, for a pair of stored (hence stationary) signals. In a time series instantaneous ... More

Some intricacies of the momentum operator in quantum mechanicsJun 06 2007Sep 23 2008In quantum mechanics textbooks the momentum operator is defined in the Cartesian coordinates and rarely the form of the momentum operator in spherical polar coordinates is discussed. Consequently one always generalizes the Cartesian prescription to other ... More

Ultra-low Energy, High-Performance Dynamic Resistive Threshold LogicAug 08 2013We propose dynamic resistive threshold-logic (DRTL) design based on non-volatile resistive memory. A threshold logic gate (TLG) performs summation of multiple inputs multiplied by a fixed set of weights and compares the sum with a threshold. DRTL employs ... More

Ultra Low Power Associative Computing with Spin Neurons and Resistive Crossbar MemoryApr 08 2013Emerging resistive-crossbar memory (RCM) technology can be promising for computationally-expensive analog pattern-matching tasks. However, the use of CMOS analog-circuits with RCM would result in large power-consumption and poor scalability, thereby eschewing ... More

A Photonic In-Memory Computing primitive for Spiking Neural Networks using Phase-Change MaterialsAug 03 2018Oct 24 2018Spiking Neural Networks (SNNs) offer an event-driven and more biologically realistic alternative to standard Artificial Neural Networks based on analog information processing. This can potentially enable energy-efficient hardware implementations of neuromorphic ... More

STDP Based Pruning of Connections and Weight Quantization in Spiking Neural Networks for Energy Efficient RecognitionOct 12 2017Spiking Neural Networks (SNNs) with a large number of weights and varied weight distribution can be difficult to implement in emerging in-memory computing hardware due to the limitations on crossbar size (implementing dot product), the constrained number ... More

Modeling and Analysis of Loading Effect in Leakage of Nano-Scaled Bulk-CMOS Logic CircuitsOct 25 2007In nanometer scaled CMOS devices significant increase in the subthreshold, the gate and the reverse biased junction band-to-band-tunneling (BTBT) leakage, results in the large increase of total leakage power in a logic circuit. Leakage components interact ... More

Localization of Dirac-like excitations in graphene in the presence of smooth inhomogeneous magnetic fieldsAug 09 2011Dec 14 2011The present article discusses magnetic confinement of the Dirac excitations in graphene in presence of inhomogeneous magnetic fields. In the first case a magnetic field directed along the z axis whose magnitude is proportional to $1/r$ is chosen. In the ... More

Design of a Low Voltage Analog-to-Digital Converter using Voltage Controlled Stochastic Switching of Low Barrier NanomagnetsMar 04 2018May 23 2018The inherent stochasticity in many nano-scale devices makes them prospective candidates for low-power computations. Such devices have been demonstrated to exhibit probabilistic switching between two stable states to achieve stochastic behavior. Recently, ... More

SPARE: Spiking Networks Acceleration Using CMOS ROM-Embedded RAM as an In-Memory-Computation PrimitiveNov 20 2017Jul 31 2018Despite huge success of artificial intelligence, hardware systems running these algorithms consume orders of magnitude higher energy compared to the human brain, mainly due to heavy data movements between the memory unit and the computation cores. Spiking ... More

Polaron master equation theory of pulse driven phonon-assisted population inversion and single photon emission from quantum dot excitonsDec 24 2015Jan 07 2016We introduce an intuitive and semi-analytical polaron master equation approach to model pulse-driven population inversion and emitted single photons from a quantum dot exciton. The master equation theory allows one to identify important phonon-induced ... More

Hybrid Spintronic-CMOS Spiking Neural Network With On-Chip Learning: Devices, Circuits and SystemsOct 01 2015Nov 13 2015Over the past decade Spiking Neural Networks (SNN) have emerged as one of the popular architectures to emulate the brain. In SNN, information is temporally encoded and communication between neurons is accomplished by means of spikes. In such networks, ... More

Hybrid Spintronic-CMOS Spiking Neural Network With On-Chip Learning: Devices, Circuits and SystemsOct 01 2015Nov 04 2016Over the past decade Spiking Neural Networks (SNN) have emerged as one of the popular architectures to emulate the brain. In SNN, information is temporally encoded and communication between neurons is accomplished by means of spikes. In such networks, ... More

Incremental Learning in Deep Convolutional Neural Networks Using Partial Network SharingDec 07 2017Sep 17 2018Deep convolutional neural network (DCNN) based supervised learning is a widely practiced approach for large-scale image classification. However, retraining these large networks to accommodate new, previously unseen data demands high computational time ... More

Conditional Deep Learning for Energy-Efficient and Enhanced Pattern RecognitionSep 29 2015Jan 28 2016Deep learning neural networks have emerged as one of the most powerful classification tools for vision related applications. However, the computational and energy requirements associated with such deep nets can be quite high, and hence their energy-efficient ... More

DSTT-MRAM: Differential Spin Hall MRAM for On-chip MemoriesMay 17 2013A new device structure for spin transfer torque based magnetic random access memory is proposed for on-chip memory applications. Our device structure exploits spin Hall effect to create a differential memory cell that exhibits fast and energy-efficient ... More

Enabling Spike-based Backpropagation in State-of-the-art Deep Neural Network ArchitecturesMar 15 2019Spiking Neural Networks (SNNs) has recently emerged as a prominent neural computing paradigm. However, the typical shallow spiking network architectures have limited capacity for expressing complex representations, while training a very deep spiking network ... More

Discretization based Solutions for Secure Machine Learning against Adversarial AttacksFeb 08 2019Feb 11 2019Adversarial examples are perturbed inputs that are designed (from a deep learning network's (DLN) parameter gradients) to mislead the DLN during test time. Intuitively, constraining the dimensionality of inputs or parameters of a network reduces the 'space' ... More

Convolutional Spike Timing Dependent Plasticity based Feature Learning in Spiking Neural NetworksMar 10 2017Mar 20 2017Brain-inspired learning models attempt to mimic the cortical architecture and computations performed in the neurons and synapses constituting the human brain to achieve its efficiency in cognitive tasks. In this work, we present convolutional spike timing ... More

Design and Synthesis of Ultra Low Energy Spin-Memristor Threshold LogicFeb 10 2014A threshold logic gate (TLG) performs weighted sum of multiple inputs and compares the sum with a threshold. We propose Spin-Memeristor Threshold Logic (SMTL) gates, which employ memristive cross-bar array (MCA) to perform current-mode summation of binary ... More

Improved current saturation and shifted switching threshold voltage in In2O3 nanowire based, fully transparent NMOS inverters via femtosecond laser annealingJul 07 2010Transistors based on various types of non-silicon nanowires have shown great potential for a variety of applications, especially for those require transparency and low-temperature substrates. However, critical requirements for circuit functionality such ... More

Discretization based Solutions for Secure Machine Learning against Adversarial AttacksFeb 08 2019Adversarial examples are perturbed inputs that are designed (from a deep learning network's (DLN) parameter gradients) to mislead the DLN during test time. Intuitively, constraining the dimensionality of inputs or parameters of a network reduces the 'space' ... More

A Low Effort Approach to Structured CNN Design Using PCADec 15 2018Deep learning models hold state of the art performance in many fields, yet their design is still based on heuristics or grid search methods. This work proposes a method to analyze a trained network and deduce a redundancy-free, compressed architecture ... More

Gabor Filter Assisted Energy Efficient Fast Learning Convolutional Neural NetworksMay 12 2017Convolutional Neural Networks (CNN) are being increasingly used in computer vision for a wide range of classification and recognition problems. However, training these large networks demands high computational time and energy requirements; hence, their ... More

Multiple alignment of structures using center of proteinsDec 28 2014In this paper we report on an algorithm for aligning multiple protein structures. The algorithm has been tested on a variety of inputs and it performs well in comparison to well-known algorithms for this problem.

Spin Neurons: A Possible Path to Energy-Efficient Neuromorphic ComputersSep 12 2013Recent years have witnessed growing interest in the field of brain-inspired computing based on neural-network architectures. In order to translate the related algorithmic models into powerful, yet energy-efficient cognitive-computing hardware, computing-devices ... More

TraNNsformer: Neural network transformation for memristive crossbar based neuromorphic system designAug 26 2017Mar 04 2018Implementation of Neuromorphic Systems using post Complementary Metal-Oxide-Semiconductor (CMOS) technology based Memristive Crossbar Array (MCA) has emerged as a promising solution to enable low-power acceleration of neural networks. However, the recent ... More

Energy-Efficient Memories using Magneto-Electric Switching of FerromagnetsJan 27 2017Voltage driven magneto-electric (ME) switching of ferro-magnets has shown potential for future low-energy spintronic memories. In this paper, we first analyze two different ME devices viz. ME-MTJ and ME-XNOR device with respect to writability, readability ... More

A Low Effort Approach to Structured CNN Design Using PCADec 15 2018Mar 12 2019Deep learning models hold state of the art performance in many fields, yet their design is still based on heuristics or grid search methods. This work proposes a method to analyze a trained network and deduce an optimized, compressed architecture that ... More

Proposal for an All-Spin Artificial Neural Network: Emulating Neural and Synaptic Functionalities Through Domain Wall Motion in FerromagnetsOct 02 2015Feb 03 2016Non-Boolean computing based on emerging post-CMOS technologies can potentially pave the way for low-power neural computing platforms. However, existing work on such emerging neuromorphic architectures have either focused on solely mimicking the neuron, ... More

Boolean and Non-Boolean Computation With Spin DevicesApr 19 2013Aug 22 2013Recently several device and circuit design techniques have been explored for applying nano-magnets and spin torque devices like spin valves and domain wall magnets in computational hardware. However, most of them have been focused on digital logic, and, ... More

Ultra-low Energy, High Performance and Programmable Magnetic Threshold LogicAug 08 2013We propose magnetic threshold-logic (MTL) design based on non-volatile spin-torque switches. A threshold logic gate (TLG) performs summation of multiple inputs multiplied by a fixed set of weights and compares the sum with a threshold. MTL employs resistive ... More

Efficient Hybrid Network Architectures for Extremely Quantized Neural Networks Enabling Intelligence at the EdgeFeb 01 2019The recent advent of `Internet of Things' (IOT) has increased the demand for enabling AI-based edge computing. This has necessitated the search for efficient implementations of neural networks in terms of both computations and storage. Although extreme ... More

Proposal for a Leaky-Integrate-Fire Spiking Neuron based on Magneto-Electric Switching of Ferro-magnetsSep 29 2016The efficiency of the human brain in performing classification tasks has attracted considerable research interest in brain-inspired neuromorphic computing. Hardware implementations of a neuromorphic system aims to mimic the computations in the brain through ... More

System Software: Concepts and ApproachMay 07 2014In software industry a large number of projects continue to fail due to non technical issue such as communication gap,requirements and poor executive. The authors identify the reasons for which are available for software development life cycles fall short ... More

On the global visibility of a singularity in spherically symmetric gravitational collapseJun 12 2014We revisit the gravitational collapse of spherically symmetric Lema\^itre - Tolman - Bondi (LTB) dust models. A sufficient condition for global visibility of singularity is given. This condition also allows us to extend the condition of local visibility ... More

Length contraction in Very Special RelativityMay 21 2011Glashow and Cohen claim that many results of special theory of relativity (SR) like time dilation, relativistic velocity addition, etc, can be explained by using certain proper subgroups, of the Lorentz group, which collectively form the main body of ... More

A Geometric Analysis of Time Series Leading to Information Encoding and a New Entropy MeasureOct 13 2018A time series is uniquely represented by its geometric shape, which also carries information. A time series can be modelled as the trajectory of a particle moving in a force field with one degree of freedom. The force acting on the particle shapes the ... More

Performance Analysis of CSMA/CA based Medium Access in Full Duplex Wireless CommunicationsDec 13 2015Full duplex communication promises a paradigm shift in wireless networks by allowing simultaneous packet transmission and reception within the same channel. While recent prototypes indicate the feasibility of this concept, there is a lack of rigorous ... More

I-Min: An Intelligent Fermat Point Based Energy Efficient Geographic Packet Forwarding Technique for Wireless Sensor and Ad Hoc NetworksJun 18 2010Energy consumption and delay incurred in packet delivery are the two important metrics for measuring the performance of geographic routing protocols for Wireless Adhoc and Sensor Networks (WASN). A protocol capable of ensuring both lesser energy consumption ... More

Reality of linear and angular momentum expectation values in bound statesApr 03 2007In quantum mechanics textbooks the momentum operator is defined in the Cartesian coordinates and rarely the form of the momentum operator in spherical polar coordinates is discussed. Consequently one always generalizes the Cartesian prescription to other ... More

Resonance fluorescence spectra from semiconductor quantum dots coupled to slow-light photonic crystal waveguidesMar 09 2016Using a polaron master equation approach we investigate the resonance fluorescence spectra from coherently driven quantum dots (QDs) coupled to an acoustic phonon bath and a photonic crystal waveguide with a rich local density of photon states (LDOS). ... More

Going Deeper in Spiking Neural Networks: VGG and Residual ArchitecturesFeb 07 2018Over the past few years, Spiking Neural Networks (SNNs) have become popular as a possible pathway to enable low-power event-driven neuromorphic hardware. However, their application in machine learning have largely been limited to very shallow neural network ... More

FALCON: Feature Driven Selective Classification for Energy-Efficient Image RecognitionSep 12 2016Mar 08 2017Machine-learning algorithms have shown outstanding image recognition or classification performance for computer vision applications. However, the compute and energy requirement for implementing such classifier models for large-scale problems is quite ... More

Spin-Orbit Torque Induced Spike-Timing Dependent PlasticityDec 19 2014Nanoelectronic devices that mimic the functionality of synapses are a crucial requirement for performing cortical simulations of the brain. In this work we propose a ferromagnet-heavy metal heterostructure that employs spin-orbit torque to implement Spike-Timing ... More

Stochastic Spin-Orbit Torque Devices as Elements for Bayesian InferenceJul 15 2017Sep 11 2017Probabilistic inference from real-time input data is becoming increasingly popular and may be one of the potential pathways at enabling cognitive intelligence. As a matter of fact, preliminary research has revealed that stochastic functionalities also ... More

Multiplier-less Artificial Neurons Exploiting Error Resiliency for Energy-Efficient Neural ComputingFeb 27 2016Large-scale artificial neural networks have shown significant promise in addressing a wide range of classification and recognition applications. However, their large computational requirements stretch the capabilities of computing platforms. The fundamental ... More

Going Deeper in Spiking Neural Networks: VGG and Residual ArchitecturesFeb 07 2018Feb 19 2019Over the past few years, Spiking Neural Networks (SNNs) have become popular as a possible pathway to enable low-power event-driven neuromorphic hardware. However, their application in machine learning have largely been limited to very shallow neural network ... More

Buttiker Probe Based Modeling of TDDB: Application to Dielectric Breakdown in MTJs and MOS DevicesAug 13 2016Dielectric layers are gradually being down-scaled in different electronic devices like MOSFETs and Magnetic Tunnel Junctions (MTJ) with shrinking device sizes. As a result, time dependent dielectric breakdown (TDDB) has become a major issue in such devices. ... More

FALCON: Feature Driven Selective Classification for Energy-Efficient Image RecognitionSep 12 2016Machine-learning algorithms have shown outstanding image recognition or classification performance for computer vision applications. However, the compute and energy requirement for implementing such classifier models for large-scale problems is quite ... More

Statistical Modeling of Pipeline Delay and Design of Pipeline under Process Variation to Enhance Yield in sub-100nm TechnologiesOct 25 2007Operating frequency of a pipelined circuit is determined by the delay of the slowest pipeline stage. However, under statistical delay variation in sub-100nm technology regime, the slowest stage is not readily identifiable and the estimation of the pipeline ... More

An All-Memristor Deep Spiking Neural Computing System: A Step Towards Realizing the Low Power,Stochastic BrainDec 05 2017Apr 13 2018Deep 'Analog Artificial Neural Networks' (ANNs) perform complex classification problems with remarkably high accuracy. However, they rely on humongous amount of power to perform the calculations, veiling the accuracy benefits. The biological brain on ... More

Spin-Torque Sensors for Energy Efficient High Speed Long InterconnectsDec 02 2015In this paper, we propose a Spin-Torque (ST) based sensing scheme that can enable energy efficient multi-bit long distance interconnect architectures. Current-mode interconnects have recently been proposed to overcome the performance degradations associated ... More

Thermoelectric Spin-Transfer Torque MRAM with Sub-Nanosecond Bi-Directional Writing using Magnonic CurrentAug 11 2011A new genre of Spin-Transfer Torque (STT) MRAM is proposed, in which bi-directional writing is achieved using thermoelectrically controlled magnonic current as an alternative to conventional electric current. The device uses a magnetic tunnel junction ... More

Exploring Boolean and Non-Boolean Computing Applications of Spin Torque DevicesAug 13 2013Aug 16 2013In this paper we discuss the potential of emerging spintorque devices for computing applications. Recent proposals for spinbased computing schemes may be differentiated as all-spin vs. hybrid, programmable vs. fixed, and, Boolean vs. non-Boolean. All ... More

Probabilistic Deep Spiking Neural Systems Enabled by Magnetic Tunnel JunctionMay 15 2016Deep Spiking Neural Networks are becoming increasingly powerful tools for cognitive computing platforms. However, most of the existing literature on such computing models are developed with limited insights on the underlying hardware implementation, resulting ... More

An Energy-Efficient Mixed-Signal Neuron for Inherently Error-Resilient Neuromorphic SystemsOct 24 2017This work presents the design and analysis of a mixed-signal neuron (MS-N) for convolutional neural networks (CNN) and compares its performance with a digital neuron (Dig-N) in terms of operating frequency, power and noise. The circuit-level implementation ... More

Buttiker Probe Based Modeling of TDDB: Application to Dielectric Breakdown in MTJs and MOS DevicesAug 13 2016Feb 24 2017Dielectric layers are gradually being down-scaled in different electronic devices like MOSFETs and Magnetic Tunnel Junctions (MTJ) with shrinking device sizes. As a result, time dependent dielectric breakdown (TDDB) has become a major issue in such devices. ... More

Voltage-Driven Domain-Wall Motion based Neuro-Synaptic Devices for Dynamic On-line LearningMay 19 2017Nov 23 2017Conventional von-Neumann computing models have achieved remarkable feats for the past few decades. However, they fail to deliver the required efficiency for certain basic tasks like image and speech recognition when compared to biological systems. As ... More

Magnetic Tunnel Junction Mimics Stochastic Cortical Spiking NeuronsOct 01 2015Jul 23 2016Brain-inspired computing architectures attempt to mimic the computations performed in the neurons and the synapses in the human brain in order to achieve its efficiency in learning and cognitive tasks. In this work, we demonstrate the mapping of the probabilistic ... More

ASP: Learning to Forget with Adaptive Synaptic Plasticity in Spiking Neural NetworksMar 22 2017Jun 08 2018A fundamental feature of learning in animals is the "ability to forget" that allows an organism to perceive, model and make decisions from disparate streams of information and adapt to changing environments. Against this backdrop, we present a novel unsupervised ... More

Energy-Efficient Object Detection using Semantic DecompositionSep 29 2015Sep 20 2016Machine-learning algorithms offer immense possibilities in the development of several cognitive applications. In fact, large scale machine-learning classifiers now represent the state-of-the-art in a wide range of object detection/classification problems. ... More

RESPARC: A Reconfigurable and Energy-Efficient Architecture with Memristive Crossbars for Deep Spiking Neural NetworksFeb 20 2017Neuromorphic computing using post-CMOS technologies is gaining immense popularity due to its promising abilities to address the memory and power bottlenecks in von-Neumann computing systems. In this paper, we propose RESPARC - a reconfigurable and energy ... More

Ultra Low Energy Analog Image Processing Using Spin NeuronsJun 12 2012Apr 08 2013In this work we present an ultra low energy, 'on-sensor' image processing architecture, based on cellular array of spin based neurons. The 'neuron' constitutes of a lateral spin valve (LSV) with multiple input magnets, connected to an output magnet, using ... More

Energy-Efficient and Robust Associative Computing with Electrically Coupled Dual Pillar Spin-Torque OscillatorsSep 12 2013Dynamics of coupled spin-torque oscillators can be exploited for non-Boolean information processing. However, the feasibility of coupling large number of STOs with energy-efficiency and sufficient robustness towards parameter-variation and thermal-noise, ... More

Toward Fast Neural Computing using All-Photonic Phase Change Spiking NeuronsApr 01 2018Aug 28 2018The rapid growth of brain-inspired computing coupled with the inefficiencies in the CMOS implementations of neuromrphic systems has led to intense exploration of efficient hardware implementations of the functional units of the brain, namely, neurons ... More

RxNN: A Framework for Evaluating Deep Neural Networks on Resistive CrossbarsAug 31 2018Jan 18 2019Resistive crossbars have emerged as promising building blocks for realizing DNNs due to their ability to compactly and efficiently realize the dominant DNN computational kernel, viz., vector-matrix multiplication. However, a key challenge with resistive ... More

Software-Defined Network Controlled Switching between Millimeter Wave and Terahertz Small CellsFeb 09 2017Small cells are a cost-effective way to reliably expand network coverage and provide significantly increased capacity for end users. The ultra-high bandwidth available at millimeter (mmWave) and Terahertz (THz) frequencies can effectively realize short-range ... More