It does have something to do with RAM and how many other resources your pc is using at the same time, but the speed is something that the programmers are working on for future versions. It also depends on the recordings. Some people do not trim their sound files, so there ends up being silence before and after the pronunciation, which makes it flash slower.
Did you read my post, Joha? It is not only dependent on the RAM. Do you have Vista or XP? I recommend having 3gb if you have Vista.