Sumeet Kumar Gupta
YOU?
Author Swipe
View article: InterAxNN: Reconfigurable and Approximate in-Memory Processing Accelerator for Ultra-Low-Power Binary Neural Network Inference in Intermittently Powered Systems
InterAxNN: Reconfigurable and Approximate in-Memory Processing Accelerator for Ultra-Low-Power Binary Neural Network Inference in Intermittently Powered Systems Open
In this work, we propose InterAxNN , an energy-aware approximate hardware architecture to perform vector-matrix multiplications in the binary precision regime for energy-constrained intermittently powered systems (IPS). In contrast to exis…
View article: Reverse Designing Ferroelectric Capacitors with Machine Learning-based Compact Modeling
Reverse Designing Ferroelectric Capacitors with Machine Learning-based Compact Modeling Open
Machine learning-based compact models provide a rapid and efficient approach for estimating device behavior across multiple input parameter variations. In this study, we introduce two reverse-design algorithms that utilize these compact mo…
View article: ReTern: Exploiting Natural Redundancy and Sign Transformations for Enhanced Fault Tolerance in Compute-in-Memory based Ternary LLMs
ReTern: Exploiting Natural Redundancy and Sign Transformations for Enhanced Fault Tolerance in Compute-in-Memory based Ternary LLMs Open
Ternary large language models (LLMs), which utilize ternary precision weights and 8-bit activations, have demonstrated competitive performance while significantly reducing the high computational and memory requirements of full-precision LL…
View article: Models for Spatially Resolved Conductivity of Rectangular Interconnects with Integrated Effect of Surface And Grain Boundary Scattering
Models for Spatially Resolved Conductivity of Rectangular Interconnects with Integrated Effect of Surface And Grain Boundary Scattering Open
Surface scattering and grain boundary scattering are two prominent mechanisms dictating the conductivity of interconnects and are traditionally modeled using the Fuchs-Sondheimer (FS) and Mayadas-Shatzkes (MS) theories, respectively. In ad…
View article: Small-signal capacitance in ferroelectric hafnium zirconium oxide: mechanisms and physical insights
Small-signal capacitance in ferroelectric hafnium zirconium oxide: mechanisms and physical insights Open
Phase-field simulation study of the mechanisms governing small-signal capacitance in ferroelectric hafnium zirconium oxide and correlation to polarization domain configurations and switching mechanisms.
View article: 1.58-b FeFET-Based Ternary Neural Networks: Achieving Robust Compute-In-Memory With Weight-Input Transformations
1.58-b FeFET-Based Ternary Neural Networks: Achieving Robust Compute-In-Memory With Weight-Input Transformations Open
Ternary weight neural networks (TWNs), with weights quantized to three states (−1, 0, and 1), have emerged as promising solutions for resource-constrained edge artificial intelligence (AI) platforms due to their high energy efficiency with…
View article: Oxygen Vacancy-Induced Monoclinic Dead Layers in Ferroelectric $Hf_xZr_{1-x}O_2$ With Metal Electrodes
Oxygen Vacancy-Induced Monoclinic Dead Layers in Ferroelectric $Hf_xZr_{1-x}O_2$ With Metal Electrodes Open
In this work, we analyze dead layer comprising non-polar monoclinic (m) phase in $Hf_xZr_{1-x}O_2$ (HZO)-based ferroelectric (FE) material using first principles analysis. We show that with widely used tungsten (W) metal electrode, the spa…
View article: BinSparX: Sparsified Binary Neural Networks for Reduced Hardware Non-Idealities in Xbar Arrays
BinSparX: Sparsified Binary Neural Networks for Reduced Hardware Non-Idealities in Xbar Arrays Open
Compute-in-memory (CiM)-based binary neural network (CiM-BNN) accelerators marry the benefits of CiM and ultra-low precision quantization, making them highly suitable for edge computing. However, CiM-enabled crossbar (Xbar) arrays are plag…
View article: Experimental Investigation of Variations in Polycrystalline Hf0.5Zr0.5O2 (HZO)-based MFIM
Experimental Investigation of Variations in Polycrystalline Hf0.5Zr0.5O2 (HZO)-based MFIM Open
Device-to-device variations in ferroelectric (FE) hafnium oxide (HfO2)-based devices pose a crucial challenge that limits the otherwise promising capabilities of this technology. Although previous simulation-based studies have identified p…
View article: The Impact of TaS$_{2}$-Augmented Interconnects on Circuit Performance: A Temperature-Dependent Analysis
The Impact of TaS$_{2}$-Augmented Interconnects on Circuit Performance: A Temperature-Dependent Analysis Open
Monolayer TaS$_{2}$ is being explored as a future liner/barrier to circumvent the scalability issues of the state-of-the-art interconnects. However, its large vertical resistivity poses some concerns and mandates a comprehensive circuit an…
View article: Comparative Evaluation of Memory Technologies for Synaptic Crossbar Arrays- Part 2: Design Knobs and DNN Accuracy Trends
Comparative Evaluation of Memory Technologies for Synaptic Crossbar Arrays- Part 2: Design Knobs and DNN Accuracy Trends Open
Crossbar memory arrays have been touted as the workhorse of in-memory computing (IMC)-based acceleration of Deep Neural Networks (DNNs), but the associated hardware non-idealities limit their efficacy. To address this, cross-layer design s…
View article: Memory Faults in Activation-sparse Quantized Deep Neural Networks: Analysis and Mitigation using Sharpness-aware Training
Memory Faults in Activation-sparse Quantized Deep Neural Networks: Analysis and Mitigation using Sharpness-aware Training Open
Improving the hardware efficiency of deep neural network (DNN) accelerators with techniques such as quantization and sparsity enhancement have shown an immense promise. However, their inference accuracy in non-ideal real-world settings (su…
View article: Formation and energetics of head-to-head and tail-to-tail domain walls in hafnium zirconium oxide
Formation and energetics of head-to-head and tail-to-tail domain walls in hafnium zirconium oxide Open
View article: Spatially Resolved Conductivity of Rectangular Interconnects considering Surface Scattering -- Part II: Circuit-Compatible Modeling
Spatially Resolved Conductivity of Rectangular Interconnects considering Surface Scattering -- Part II: Circuit-Compatible Modeling Open
Interconnect conductivity modeling is a critical aspect for modern chip design. Surface scattering -- an important scattering mechanism in scaled interconnects is usually captured using Fuchs-Sondheimer (FS) model which offers the average …
View article: Spatially Resolved Conductivity of Rectangular Interconnects considering Surface Scattering -- Part I: Physical Modeling
Spatially Resolved Conductivity of Rectangular Interconnects considering Surface Scattering -- Part I: Physical Modeling Open
Accurate modeling of interconnect conductivity is important for performance evaluation of chips in advanced technologies. Surface scattering in interconnects is usually treated by using Fuchs-Sondheimer (FS) approach. While the FS model of…
View article: Ferroelectric Transistor-Based Synaptic Crossbar Arrays: The Impact of Ferroelectric Thickness and Device-Circuit Interactions
Ferroelectric Transistor-Based Synaptic Crossbar Arrays: The Impact of Ferroelectric Thickness and Device-Circuit Interactions Open
Ferroelectric transistors (FeFETs)-based crossbar arrays have shown immense promise for computing-in-memory (CiM) architectures targeted for neural accelerator designs. Offering CMOS compatibility, nonvolatility, compact bit cell, and CiM-…
View article: Interaxnn: A Reconfigurable and Approximate In-Memory Processing Accelerator for Ultra-Low-Power Binary Neural Network Inference in Intermittently Powered Systems
Interaxnn: A Reconfigurable and Approximate In-Memory Processing Accelerator for Ultra-Low-Power Binary Neural Network Inference in Intermittently Powered Systems Open
View article: Special Topic on Physics-Based Modeling and Simulation of Materials, Devices, and Circuits of Beyond-CMOS Logic and Memory Technologies for Energy-Efficient Computing
Special Topic on Physics-Based Modeling and Simulation of Materials, Devices, and Circuits of Beyond-CMOS Logic and Memory Technologies for Energy-Efficient Computing Open
Standard complementary metal–oxide–semiconductor (CMOS) technology and its advanced flavors in the form of FinFETs have propelled the electronic industry to its extraordinary success. While the CMOS technology may continue to deliver its r…
View article: Direction‐Dependent Lateral Domain Walls in Ferroelectric Hafnium Zirconium Oxide and their Gradient Energy Coefficients: A First‐Principles Study
Direction‐Dependent Lateral Domain Walls in Ferroelectric Hafnium Zirconium Oxide and their Gradient Energy Coefficients: A First‐Principles Study Open
To understand and harness the physical mechanisms of ferroelectric hafnium zirconium oxide (HZO)‐based devices, there is a need for clear understanding of domain interactions, their dynamics, negative capacitance effects, and other multido…
View article: XNOR-VSH: A Valley-Spin Hall Effect-Based Compact and Energy-Efficient Synaptic Crossbar Array for Binary Neural Networks
XNOR-VSH: A Valley-Spin Hall Effect-Based Compact and Energy-Efficient Synaptic Crossbar Array for Binary Neural Networks Open
Binary neural networks (BNNs) have shown an immense promise for resource-constrained edge artificial intelligence (AI) platforms. However, prior designs typically either require two bit-cells to encode signed weights leading to an area ove…
View article: Reimagining Sense Amplifiers: Harnessing Phase Transition Materials for Current and Voltage Sensing
Reimagining Sense Amplifiers: Harnessing Phase Transition Materials for Current and Voltage Sensing Open
Energy-efficient sense amplifier (SA) circuits are essential for reliable detection of stored memory states in emerging memory systems. In this work, we present four novel sense amplifier (SA) topologies based on phase transition material …
View article: Harnessing Unipolar Threshold Switches for Enhanced Rectification
Harnessing Unipolar Threshold Switches for Enhanced Rectification Open
Phase transition materials (PTM) have drawn significant attention in recent years due to their abrupt threshold switching characteristics and hysteretic behavior. Augmentation of the PTM with a transistor has been shown to provide enhanced…
View article: Comparative Evaluation of Memory Technologies for Synaptic Crossbar Arrays -- Part I: Robustness-driven Device-Circuit Co-Design and System Implications
Comparative Evaluation of Memory Technologies for Synaptic Crossbar Arrays -- Part I: Robustness-driven Device-Circuit Co-Design and System Implications Open
In-memory computing (IMC) utilizing synaptic crossbar arrays is promising for energy-efficient deep neural network (DNN) accelerators. Various technologies (CMOS and post-CMOS) have been explored as synaptic device candidates, each with it…
View article: XNOR-VSH: A Valley-Spin Hall Effect-based Compact and Energy-Efficient Synaptic Crossbar Array for Binary Neural Networks
XNOR-VSH: A Valley-Spin Hall Effect-based Compact and Energy-Efficient Synaptic Crossbar Array for Binary Neural Networks Open
Binary neural networks (BNNs) have shown an immense promise for resource-constrained edge artificial intelligence (AI) platforms as their binarized weights and inputs can significantly reduce the compute, storage and communication costs. S…
View article: Head-to-Head and Tail-to-Tail Domain Wall in Hafnium Zirconium Oxide: A First Principles Analysis of Domain Wall Formation and Energetics
Head-to-Head and Tail-to-Tail Domain Wall in Hafnium Zirconium Oxide: A First Principles Analysis of Domain Wall Formation and Energetics Open
180° domains walls (DWs) of Head-to-Head/Tail-to-Tail (H-H/T-T) type in ferroelectric (FE) materials are of immense interest for a comprehensive understanding of the FE attributes as well as harnessing them for new applications. Our first …
View article: Phase-field Simulations of Polarization Variations in Polycrystalline Hf0.5Zr0.5O2 based MFIM: Voltage-Dependence and Dynamics
Phase-field Simulations of Polarization Variations in Polycrystalline Hf0.5Zr0.5O2 based MFIM: Voltage-Dependence and Dynamics Open
In this work, we investigate the device-to-device variations in remanent polarization of Hafnium-Zirconium-Oxide based Metal-Ferroelectric-Insulator-Metal (MFIM) stacks. We consider the effects of polycrystallinity in conjunction with mult…
View article: FlatENN: Train Flat for Enhanced Fault Tolerance of Quantized Deep Neural Networks
FlatENN: Train Flat for Enhanced Fault Tolerance of Quantized Deep Neural Networks Open
Model compression via quantization and sparsity enhancement has gained an immense interest to enable the deployment of deep neural networks (DNNs) in resource-constrained edge environments. Although these techniques have shown promising re…
View article: Valley-Spin Hall Effect-Based Nonvolatile Memory With Exchange-Coupling-Enabled Electrical Isolation of Read and Write Paths
Valley-Spin Hall Effect-Based Nonvolatile Memory With Exchange-Coupling-Enabled Electrical Isolation of Read and Write Paths Open
Valley-spin hall (VSH) effect in monolayer WSe2 has been shown to exhibit highly beneficial features for nonvolatile memory (NVM) design. Key advantages of VSH-based magnetic random access memory (VSH-MRAM) over spin orbit torque (SOT)-MRA…
View article: Negative Capacitors and Applications
Negative Capacitors and Applications Open
View article: Valley-Spin Hall Effect-based Nonvolatile Memory with Exchange-Coupling-Enabled Electrical Isolation of Read and Write Paths
Valley-Spin Hall Effect-based Nonvolatile Memory with Exchange-Coupling-Enabled Electrical Isolation of Read and Write Paths Open
Valley-spin hall (VSH) effect in monolayer WSe2 has been shown to exhibit highly beneficial features for nonvolatile memory (NVM) design. Key advantages of VSH-based magnetic random-access memory (VSH-MRAM) over spin orbit torque (SOT)-MRA…