In a typical pixel detector using submicron CMOS technology one is now able to integrate in each microscopic pixel a traditional nuclear electronics readout chain, including ADC and data storage. However, the technology is approaching 10nm dimensions for active elements, and one can dream of more complicated integration. Will this allow to create new systems for scientific studies ? Probably one needs to integrate also adjacent technological developments in miniaturization of actuators and microfluidic cooling. The current approach to nanoelectronics is based on mass production with very expensive lithographic masks. For scientific applications, alternatives are becoming very much needed.