Simulating Noisy Diamond Quantum Computers
Abstract
Quantum computers have enormous promise in areas as diverse as medicine, finance, defence and information security. NV-centres in diamond are a promising platform for quantum computing due to their long coherence times at room temperature. However, current quantum computers are characterised by their noisy operation and do not provide any quantum advantage. Classical simulations can be used to provide feedback on various sources of noise to guide future engineering designs. Simulations can also be used to benchmark real quantum computers, and develop new quantum algorithms. Unfortunately, current noisy simulators are phenomenological and limited in scale. Hence they often poorly represent the general behaviour of real quantum computers, limiting their utility. In this thesis I work towards developing an accurate, efficient simulator of noisy diamond quantum computers, with results generalisable to other architectures. I first compare the performance of two state of the art simulation methods to identify which is the most efficient. By performing runtime comparisons, I demonstrate that a leading Schrodinger simulator outperforms a leading tensor network simulator when performing high fidelity simulations, at least for fewer than 30 qubits. Next, I derive and validate a noise model based on decoherence and single-qubit control errors in diamond quantum computers. By simulating single-qubit circuits, I show that a depolarising channel produces simulations with a higher fidelity than the derived model. I attribute this to the small errors in single-qubit circuits favouring the fitted depolarising channel, which would not hold for multi-qubit circuits. I conclude that while the derived model requires further developments, it still offers immediate advantages over phenomenological models.