Research Overview

Concentrating on the development and optimization of emerging nanoelectronic devices, device modeling is a crucial aspect of our research. By employing advanced computational techniques, we aim to gain deep insights into the fundamental properties and behaviors of these novel devices. To understand and predict the electronic and magnetic behavior at the quantum level, we utilize non-equilibrium Green's function (NEGF) approaches to model spin transport phenomena in MTJs. Our research involves modeling the stochastic switching mechanisms of emerging nanoelectronic devices using the Fokker-Planck approach. This method helps us capture the probabilistic nature of switching events in devices such as resistive RAM (ReRAM) and other non-volatile memories, which is essential for optimizing their reliability and performance. Our modeling efforts are closely integrated with experimental data to validate and refine our theoretical predictions. This iterative process ensures that our models are accurate and reliable, providing valuable insights that guide experimental work. One of our primary goals is to optimize these devices for energy efficiency. By understanding the underlying mechanisms at the nanoscale, we aim to design devices that consume less power while maintaining high performance, which is critical for applications in low-power electronics and neuromorphic computing.

Mimicking neuro-synaptic behavior using emerging devices aims to replicate the complex functionality of the human brain in electronic systems. We explore spintronic devices and other novel materials to create artificial neurons and synapses, which form the building blocks of neuromorphic systems that emulate the neural structures and functions of biological brains. Our models simulate dynamic processes such as synaptic plasticity, spiking behavior, and learning mechanisms. These artificial neurons and synapses are integrated into larger neuromorphic architectures, involving the design of circuits that support spiking neural networks (SNNs) and other brain-inspired computing paradigms. We focus on optimizing these devices to minimize power consumption while maintaining high performance, making them suitable for applications in edge computing and low-power environments. Our neuromorphic systems have various applications, including pattern recognition, real-time data processing, and autonomous decision-making. By mimicking neuro-synaptic behavior, we aim to develop computing systems capable of performing complex tasks with the efficiency and adaptability of the human brain.