System Architecture and Complexity

5Generations of Systems and the System in the System

A system is never formed from nothing at all. There are always antecedents; it is part of a genealogy, which means that we can talk about generations and/or successive adaptations, those in generation N drawing lessons from the previous generations. ICT has made us familiar with this notion of generation, but it goes much further back¹. Without going all the way back to antiquity (a Greek or Roman sailing ship has all the characteristics of a system, starting with the pilot!), or if we look within the contemporary era, we can cite tool machines, now robots, generations of computers, today we have System On Chip (SoC) with circuits of 2–3 cm², which weigh a few grams and which contain 10 billion transistors, or even the stages of the electronuclear program, where we go from turbo alternator machines with a power level of 75 MW to the current groups with a power level of approximately 1400 MW (EPR can reach 1650 MW), and this can be done in five to six stages over approximately 20 years.

What we should note specifically, from a systemics point of view, is that machines and/or systems can be used to make new machines, more powerful and more reliable. This is true for the three examples cited above, but the most interesting is perhaps the computer itself, taking into account the simulation capacities that it places in the hands of the engineers and scientists that develop the systems. This is totally explicit for J. von Neumann ...