compiling always generates fast code?

[deeptech71 at gmail.com]

Do you get processor-specific code when you do buildworld and buildkernel, 
particularly on a -RELEASE?

Does building from ports always automatically optimize code (vs packages)?

Do I need to strip some files after the compilation (world, kernel, ports) to 
get the absolute maximum performance?

Any other tips to speed up executables?

THX!