L1 & L2 Cache matters most when you're doing processing routines that require a fair amount of memory usage. Having a larger cache size can keep your working set in cache, preventing a (costly) memory fetch from RAM.
I run into this with my software quite frequently. We deal with large-scale mathematics problems. In this case, having a larger cache can have a profound effect on overall speed.
However, for programming, in general, I've found that smaller L2 cache sizes in processors doesn't seem to have a huge effect on total compilation times. For programming itself, I'd say it doesn't matter a lot. When you run the program you generate, depending on the nature of the software, it can make a big difference. You might want to read up on the
theory behind CPU caching . Having a larger cache definitely helps in some circumstances, but often, it's not (IMO) worth the extra processor costs. (Some people have told me differently - that a larger cpu cache helped reduce their compilation times dramatically - but that's probably more a matter of what specifically you're compiling. My experience has always been that it has a pretty minor effect.)
Reed Copsey, Jr. -
http://reedcopsey.com