Noise-resilient quantum algorithms and software for quantum computers

Quantum computers have immense potential to solve problems of relevance for wide ranges of industrial applications, and their development forms a key part of current Quantum and Digital Strategies. However, quantum computers are inherently probabilistic, and therefore their calculations have associated uncertainty and error bars. Importantly, to achieve their potential and be adopted by users in industry, good estimates of these error bars are needed in order to provide the required trust in the results of quantum computers. Since quantum computers will allow to simulate systems not accessible on conventional computers, the problem is that it is infeasible to verify these errors by simulations on conventional computers. The objective of this project therefore is to quantify the uncertainty based on the measurable error parameters of the quantum device. Quantum metrology can determine the magnitude of each noise source for a given device, but the key contribution of this project is to link, via metrological analysis and uncertainty propagation, the contributions of different noise terms to the uncertainty of the final result. To this aim Bayesian hierarchical models for uncertainty quantification will be developed for quantum algorithm verification, and applied to quantum computing software in the fields of chemistry simulations, condensed matter physics, and machine learning, areas where quantum computers are expected to bring the largest impact in the near and medium term.