Towards Photonic Reservoir Computing: Can a single dynamical node replace a complex network?

Finding and implementing more efficient methods for information processing are key challenges in our information-driven society. The fascination for the ability of the brain to process information, and our emerging understanding of complex systems and their control open new opportunities to perform computation beyond the Turing concept. In this talk, we explain the novel neuro-inspired concept called reservoir computing. This concept relies on recurrent complex networks, requiring the interconnection of a large number of nonlinear nodes. We will discuss in how far we can conceptually and technologically reduce the concept to a single nonlinear node with delayed feedback as processing reservoir. The final aim of the European Project PHOCUS, in which IFISC plays an important role, is to implement and demonstrate this concept in photonics. We will show that major breakthroughs have been achieved, demonstrating efficient information processing by a single nonlinear node with delayed feedback.