On UNIX, process creation is done using the kernel C interface fork (some also have vfork). The fork function creates an (almost) exact copy of the parent process, and after the fork the same code runs in both parent and child. Running another program in the child is common, and this is usually done using one of the exec family of functions to replace the current program with another.

In Perl, you can use fork in exactly the same way. Usually it is used to run an asynchronous process, that is when you don't want to wait for the child to complete. In your code you are waiting for the child to complete, which rather defeats the object. You don't have to wait at once in the parent, you can defer the wait until after all your child processes have started, see the doc for waitpid.

A couple of things to watch. First, on most UNIXs, when a child process ends it wants to tell the parent its exit code, so it waits until that has been collected, usually through waitpid or by handling the SIGCHLD signal. So if you never do that then the child cannot die, and you end up with a zombie. Second, when the parent exits it will send each child process a SIGHUP signal, which, by default, will kill the child. So make sure you wait until all the children have finished.