Dissertation/Thesis Abstract

Quantum Artificial Intelligence: Leveraging Microscopic Transport Phenomena for Beyond Von Neumann Computing
by Singh, Christopher N., Ph.D., State University of New York at Binghamton, 2020, 166; 28002984
Abstract (Summary)

Efficient processing of information has revolutionized the modern world, but transcending the current state of computing requires a paradigm shift in the way we process, store, and manipulate information. Emulating biological learning capabilities in nano-scale circuitry is a solution, but endowing materials with programmable memory is a monumental task. Quantum mechanics offers a means to do so by encoding malleable information into the wave function of a material, allowing storage and processing to happen simultaneously. Thus the question becomes: can we engineer the transport properties of quantum materials, leveraging a metal-to-insulator transition to achieve memory?

This Dissertation reviews and exposes the electronic transport processes necessary to achieve logical memory at the nano-scale level, and details the subtle interplay between lattice and electronic degrees of freedom. The work contained here is threaded by the central, recurrent theme that intrinsic material properties and device function are deeply intertwined. In a situation where the material is the device, unraveling the complex synergy between electron-electron correlations and electron-lattice correlations is a fundamental step towards realization of quantum artificial intelligence, and I show through state-of-the-art computational modeling that controversial phenomena in several transition metal oxides are the key to the next revolution in computing.

Vanadium and niobium oxides are considered due to their promising metal-to-insulator transitions. The nature of endpoint phases as well as the nature of the transition under chemical and mechanical modulation is studied, and I find the 3d manifold in vanadium dioxide is highly susceptible to enhanced correlation physics, while the larger 4d manifold in niobium dioxide is not. Additionally, hafnium oxide is considered for its application in resistive random access memory architectures where the stochastic nature of disordered transport is shown to directly impact circuit design. Finally, the metal-to-insulator transition mechanism in lithium niobite is determined to be based on driven cation motion.

Indexing (document details)
Advisor: Lee, Wei-Cheng
Commitee: Piper, L. F. J., White, Bruce E., Whittingham, M. Stanley
School: State University of New York at Binghamton
Department: Physics, Applied Physics, and Astronomy
School Location: United States -- New York
Source: DAI-B 82/4(E), Dissertation Abstracts International
Source Type: DISSERTATION
Subjects: Physics, Condensed matter physics, Quantum physics
Keywords: Artificial intelligence, Conductance fluctuations, Localization, Metal to insulator transition, Novel order, Quantum transport
Publication Number: 28002984
ISBN: 9798678159687
Copyright © 2020 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest