The goal of this theoretical and computational project is to develop optimal strategies for characterization, calibration and quantum error correction of quantum computers as they scale up to the large qubit numbers required for achieving quantum advantage over conventional computers. The studentship will cover tuition fees for home students and provides a stipend with London allowance for a period of 3.5 years. The prospective PhD student will be based both in the Quantum Technologies Department at NPL (Teddington, Greater London) as well as at Imperial College London, with an approximately even time split between the two institutions over the duration of the PhD. The close proximity between NPL and Imperial will allow for a close collaboration with regular in-person meetings with all involved partners. The student will have the opportunity to engage with a number of academic and industrial collaborators.
Description of the project In the international race towards making quantum computing practically useful, both in terms of hardware and software, significant progress is being made, with the UK quantum ecosystem emerging as a key driver. Nonetheless, understanding and correcting errors due to noise in quantum computers remains one of the key challenges that must be addressed before quantum computers can reliably outperform conventional computers. Many conventional algorithms for accurately characterizing smaller-scale quantum computers require too many measurements to be applied on the emerging larger-scale devices. This project will develop new models and algorithms to tackle this challenge by combining physics informed approaches with machine learning techniques to extract maximal information about a quantum device from a minimal number of measurements. These methods will be integrated with active learning and other AI techniques to progressively accumulate the data required for optimal implementations of quantum gates and quantum error correction protocols. We aim to unlock new capabilities for practical applications of quantum algorithms in classically-hard problems, such as integer factorization and quantum simulation.
The studentship will equally be embedded in Prof. Mintert’s quantum theory research group at Imperial College London, and the Prof. Rungger’s Quantum Software and Modelling team at NPL. Prof. Mintert has a strong focus in the use of statistical machine learning to identify the optimal use of noise quantum devices, with practical demonstrations for a number of devices. Prof. Rungger’s team at NPL is developing a number of machine learning based algorithms for characterization of quantum devices, such as the recently proposed hierarchical discrete fluctuation auto-segmentation method allowing to disentangle individual sources of decoherence from the noisy output of devices.
