Bridging the gap: Predictive theory and numerics for designing solid state quantum processors

This project is associated with a full financial support package (full fees and living expenses) provided by the Quantum Motion, a young company based in London and Oxford.

The supported student would work towards a doctorate in Simon Benjamin’s theory group in Oxford. The student would be free to participate in the full range of themes that this group explores (see the generic project ‘Exploiting first-generation quantum computers’ or look at arxiv.org for the group’s recent papers). However, there is a specific research theme, described below, that is an important focus of the project and which might be expected to take at least 50% of the student’s time.

Recent experimental progress indicates that electron spins within semiconductor structures may serve as qubits for quantum computation. The exciting implication is that a variation of conventional CMOS electronics may be suitable for large scale quantum computing, bringing that reality much closer by harnessing the established silicon fabrication industry. However there remain a number of challenges to solve to realise this prospect.

Advanced quantum processor emulation tools (such as QuEST, an open source tool developed by Simon Benjamin’s Oxford group with support from Quantum Motion) have emerged for the evaluation of general qubit architectures using conventional supercomputers. It is possible to specify a set of native qubit-qubit operations and associated error rates, so capturing the basic difference between (say) ion trap processors and silicon spin processors. Thus one can investigate how well systems of different kinds, sizes and layouts would perform quantum algorithms. However, there is a still a substantial gap between the descriptions used by those emulators and the reality in the laboratory. What is needed is predictive algorithm that can take as input a detailed design of a silicon structure and provide as output a prediction of its properties in terms of qubits, gate fidelities and errors. Then by linking with tools like QuEST, we would have the capability to predict the quantum power of silicon prototype circuits. Such a capability would accelerate progress, taking bigger steps from each hardware generation to the next by focusing on the architectures that have been predicted to be the most successful.

The key focus of the project is to indeed bridge this gap. It will require the student to possess or to develop key skills including

• Competence in mathematics for physics and analytic techniques at or above the level of a physics first degree, with modules in theoretical physics an advantage
• understand of the physics of semiconductors and spins — eventually at an advanced level
• numerical skills in programming — eventually this will include using high performance conventional computers.

Concepts related to machine learning and optimisation may be useful and a prior experience here could be valuable. An interest in quantum algorithms and understanding their properties will also be important. The successful student would interact closely with both the theory team in the host group and also the experimental team in London, as well as collaborating groups in the UK and worldwide.