The need in our society for information transmission and computation is increasing every day. But technology is facing physical limits to fulfil these projected needs focusing solely on computation as we know it. Photonic information processing has characteristics that could help overcome these limitations. Implementation of novel computational techniques, like machine learning, in photonic hardware provides excellent performance, in particular in tasks that are difficult for traditional computing approaches. In this seminar, some of the activities on photonic information processing in IFISC’s Nonlinear Photonics Lab will be presented. We will introduce the experimental scheme of an all-optical reservoir computer based on a semiconductor laser with delayed optical feedback and some of its possible topologies. We will discuss their most important properties and their dependence on key parameters of the system. We show how these key properties affect the reservoir computing performance which might pave the way to adapt the all-optical reservoir computer for different tasks.