Tools for Quantum Software Engineering

Abstract

This thesis studies the multiple levels of the quantum software development stack, spanning from high-level algorithm design to fault-tolerant architectures. We begin by presenting a novel technique for preparing non-Pauli resource states directly on the surface code. These resources can be used to enact non-Clifford logical gates which are a costly but essential requirement for universal quantum computation. Alongside this result, we demonstrate the use of automated tools for benchmarking this state preparation protocol leveraging modern hardware and innovative software approaches for accelerating the generation of results.

The following chapter focuses on resource estimation tools and addresses the problem of translating between different quantum software development platforms. In this work, we produce a tool for Q# to Cirq translation that operates in conjunction with graph visualisation tools and benchmarking tools.

We use these resource estimation tools to benchmark a supremacy-type experiment run on fault-tolerant architectures. We then utilise graph theory to prove a grid structure is optimal for demonstrating quantum advantage on a processor with bounded physical resources.

We conclude this thesis with a study on minimal gate-sets, namely the Toffoli + Hadamard gate-set, and demonstrate the process of compiling an algorithm in an alternative, restricted gate-set.