Simulations of error in quantum adiabatic computations of random 2-SAT instances

Abstract

This thesis presents a series of simulations of quantum computations using the adiabatic algorithm. The goal is to explore the effect of error, using a perturbative approach that models 1-local errors to the Hamiltonian and estimates transition probabilities out of the ground state. The data show that a perturbation in the z-direction, parallel to the alignment of the computational basis, causes a greater error than one in the x-direction, and show with good confidence (X2/[nu] = 1.7) that the variation is as sin [theta] where [theta] is the angle of the error term. An attempt to explore the change in error with the number of qubits was inconclusive--there was no measurable variation.