Artificial neural networks have been studied thoroughly over the last years as an efficient ansatz to represent states of quantum systems on classical computers. Neural-network-based quantum state tomography, the reconstruction of a quantum state based on a finite amount of projective measurements, has gained great interest due to its demonstrated power and flexibility. Generative network models furthermore enable the sampling of data which emulates projective measurements of the encoded quantum state. The recent development of spiking neuromorphic hardware promises further advances in the data generation process. These mixed-signal chips, which consist of biologically inspired analog spiking neurons, allow for the implementation of generative network models and provide an accelerated sampling procedure at low energy consumption. In this talk I will demonstrate how spiking neuromorphic hardware can be used to advance neural-network-based quantum state tomography and discuss possible applications.

Download the slides for this talk.Download ( PDF, 15938.94 MB)