quantum shadows: scalable quantum-to-classical converters

The road to revolutionary quantum computers is paved with plenty of stumbling blocks. For example, large-scale quantum experiments require considerable computing power from conventional computers in order to control experimenting and process results. This explosion in resource cost impairs information transmission between quantum and conventional systems. In turn, this impedes the use of existing computer architectures and may even impact scaling towards larger and more powerful quantum computers. The project applies a consistent, unified approach that takes all of the computing technology resources (quantum and conventional computers) into consideration. In order to solve the transmission constraints, interfaces are being developed capable of converting quantum information into meaningful conventional information. These interfaces, referred to as "shadows" combine randomization and quantum-enhanced readout strategies in order to obtain an accurate, classically-generated description of an underlying quantum system. This description enables efficient and parallelizable prediction of numerous features of the quantum system. Machine learning can then identify the essential characteristics. This way it is bridging the gap between quantum-based experiments and classical machine learning. Consequently, this interdisciplinary project successfully combines methods taken from modern computer science, quantum information and artificial intelligence. The results of the project will help expand the potential of quantum computing and drive the advancement of more reliable and efficient quantum computers. The project opens up new perspectives for applications in various domains, such as simulation, optimization, and artificial intelligence