Photons play a central role in many areas of quantum information science, either as qubit themselves or to mediate interactions between long-lived matter based qubits. Techniques for (1) high-fidelity generation, (2) precise manipulation and (3) ultra-efficient detection of quantum states of light are therefore a prerequisite for virtually all quantum technologies. A quantum photonics processor is the union of these three core technologies into a single system, and, bolstered by advances in integrated photonics, promises to be a versatile platform for quantum information science. In this talk we present recent progress towards large-scale quantum photonic processors, leveraging the platform of silicon photonics which provides unrivalled performance in terms of component density, large-scale manufacturability and compatibility with CMOS electronics. We demonstrate protocols in quantum computing, quantum simulation, and classical deep learning; and propose a next generation architecture for quantum optical neural networks. These results indicate quantum photonic processors are a versatile platform for large-scale quantum technologies.
2018-11-16 (周五) 00:10