During a game of Corewar each Warrior takes it in turns to execute a single process. In the first cycle of the round, the first warrior executes an instruction. In the next cycle, the next warrior executes an instruction. Once each warrior has taken a turn, the first warrior goes again.
If no warrior executes a
spl instruction then this is all there is to it and execution continues in this fashion with each warrior taking a turn over successive cycles until the round ends.
spl instruction is executed, the executing warrior gains a new process. Now when the warrior takes a turn it will execute one process and when it next gets a turn it will execute the other process.
For example, if there are two warriors in a match, 'imp' and 'paper'. 'Imp' has a single process (A). 'Paper' has two processes (A and B). In this situation the turn order will be as follows:
Imp (A) Paper (A) Imp (A) Paper (B) Imp (A) Paper (A) etc..
Notice that although Paper has twice as many processes as Imp, it does not get any more turns. As a result, each of Paper's processes effectively execute at half the speed of Imp's single process.
Of course a big advantage of having multiple processes is resiliency, it will be harder to kill a warrior with many processes than a more fragile warrior with only a single process. The downside of multiple processes is the added complexity of multi-threaded programming.
For more details on the
spl instruction see spl.