Off-campus SJSU users: To download campus access theses, please use the following link to log into our proxy server with your SJSU library user name and PIN.

Publication Date

Spring 2018

Degree Type

Thesis - Campus Access Only

Degree Name

Master of Science (MS)

Department

Physics and Astronomy

Advisor

Ehsan Khatami

Keywords

condensed matter, Ising model, machine learning, neural network, quantum computer, spin glass

Subject Areas

Physics

Abstract

Growing interest in quantum computers and artificial intelligence has fueled several innovations in recent years with regards to machine learning. With the construction of the first experimental quantum annealers by D-Wave, attaining the significant speedup provided by a quantum computer seems within reach. However, recent tests bring into question the degree to which these quantum effects are being utilized. The presence of temperature chaos when solving for the ground-state energy of a spin-glass system and the unfavorable scaling with classical hardness suggest that the D-Wave may really be a semiclassical machine. Using a state-of-the-art parallel tempering code and by using artificial neural networks and t-distributed stochastic neighbor embedding, we search for features in spin-glass instances that cause temperature chaos, and by extension attempt to gain some insight into the performance and scaling problems of the D-Wave quantum computer. As a result, we found that temperature chaos must be caused by structures deeper than the interactions themselves.

Share

COinS