The paper considers spiking neural P systems (SN P systems) with cooperating rules where each neuron has the same number of sets of rules, labelled identically. Each set is called a component (maybe empty). At each step only one of the components can be active for the whole system, and only the rules from the active component are enabled. Each neuron with enabled rules from this active component can fire. By using 59 neurons, a small universal SN P system with two components, working in the terminating mode, is constructed for computing functions. © Springer International Publishing Switzerland 2014.