I recall a Game Developer article about the development of Quake, where John Cash cites a tightly written inner loop in the code... to clear the screen in non-textured mode (which nobody ever really uses, and contributes a minuscule amount of performance even then). By Cash's estimate, this took (Michael Abrash, the original Intel assembly optimization god) a couple days to write.

Profile before you optimize.

--
F o x t r o t U n i f o r m
Found a typo in this node? /msg me
The hell with paco, vote for Erudil!