Speaker
Description
Although quantum annealing has shown promising results, it still
struggles to outperform classical optimization algorithms. One of the
reasons for this is that the qubit connectivities in superconducting
circuits, which are one of the most promising platforms for quantum
annealing, are not complex enough. This work focuses on the analysis
of different architectures for quantum annealing. On one hand we will
study those used in existing superconducting devices, on the other
hand we will study several types of small-world and random regular
networks that are better-understood but less-fabricable. Our goal is
to study, from the perspective of spin glasses and statistical
mechanics, the hardness of the problems that can be embedded in these
architectures, and how this hardness scales as the architectures grow.
This will be a useful consideration to design future quantum annealing
architectures.
