Memory Elements for Neuromorphic Computing Using Superconducting Ferroelectric Hybrids

Memory Elements for Neuromorphic Computing Using Superconducting Ferroelectric Hybrids

PROJECT NUMBER: 84 | AUTHOR: Sean Emerson​, Physics

MENTOR: Meenakshi Singh, Physics and Geoff Brennecka, Metallurgical and Materials Engineering

ABSTRACT

The purpose of this research is to investigate the uses of ferroelectric thin films in conjunction with superconducting thin films as potential memory devices for neuromorphic computing. In doing so, we are investigating the effect of electric fields from polarized ferroelectric PZT films on the critical current of superconducting niobium thin films. By applying this research, we hope to construct a device that, when given a variety of 9 inputs, or pixels, will be able to learn patterns in the inputs and recognize them in the future. This project aims to have the learning portion of this process take place within the physical components of the device, thus displaying the device’s capabilities for non-volatile memory.

PRESENTATION

AUTHOR BIOGRAPHY

Charles Matlock is a current senior in the Metallurgical and Materials Engineering department. He has been involved in computational research for almost 2.5 years now through the Molecular Theory Group lead by Dr. Mark Eberhart and has completed hundreds of calculations through HPC@Mines. He seeks to pursue a graduate degree, focusing on computational modeling of materials, particularly metals and ceramics. His ultimate goal in life is to develop a general model of fracture that can be used in computational models to create stronger and tougher materials.