Tim,
I do this more often then not. My scenes tend to be complex. How you break it up depends on the scene. The most common way is to render background, mid-ground, foreground with the characters being in the mid-ground. I'll usually put the scene together in one project file then determine how many parts I'm going to have then save that project file to several files then start removing things. You must be careful of shadows crossing over more then one part. Actually you must be careful of many things. This of course uses alpha channels.
A whole different way, which I've never tried is to take a black box with 1/4th of it cutaway and render the upper left, upper right and so on.
Cheers,
Rusty
Ah, I see your ideas. You seem to imply that if there are either fewer parts to a render (it's split into background/midground/foreground) or entire blocks of the render are blacked out (perhaps with a black grid with sections cut out) then I can make the rendering easier on the processor/program and get what I'd need? Does the pixel count have anything do to with it?
Tim