This system behavior is described in objective terms down to its elementary functional units. Therefore, the system can be expressed as a collection of cellular automata.
You can not only describe the physical processing units as such but also there are functional automata that emerge from the activity and interactions of the former.
In general, any system behaviour can be reduced to the rules of individual cellular automata. This agrees in part with some of the advocacies of Stephen Wolfram.
But unfortunately this knowledge is useless because we can only derive those rules as the outcome of a relational computation. That is, by the time you get to be able to describe the automata you are already able to describe the system.
We can go from there to the most simple automata rules but not the inverse. We cannot directly translate system behaviours into cellular automata behaviours.
Compliments of the Author 