The motivation behind our research in dynamical systems is an implementation of brain-inspired hardware. In the first part of the talk, we discuss the emergence of complex patterns in a nonlinear delay oscillator. Those self-organized formations represent coherent-incoherent motions known as chimera states. As the number of delays is increased, we observe the appearance of new structures such as dissipative solitons. Those structures exhibit high multistability which suggests their potential application as an optical memory. In the second part of the talk, we study the delay oscillator from reservoir computing (RC) perspective, i.e. recurrent neural networks leveraging physically existing dynamical systems. The strength of the RC technique is the possibility to deal with complex time-dependent data such as sounds and chaotic time-series. Finally, we demonstrate a stand-alone delay dynamics-based reservoir computer built on top of FPGA hardware.