Yes, but it's not exactly simple.
First, you have to build a cluster. I don't know of any other method to have two machines communicate and share a workload. Building a cluster is not an easy thing to do, but that's ultimately your decision (to build it or not).
Second, the compile needs to support multiple processors. By that I mean, the application your compiling must have "pieces" of the compile process that are not dependent on each other; they can be compiled completely independently without influencing the end result. Then, you also need to use a special version of make (pmake is it's name, I believe). The point is, if you're compiling someon else's code, then there's no guarantee that it can be split. That is, unless you want to rewrite portions of their code, Makefile, or both.
My personal opinion: unless you want to build a massive compute farm, then it's probably not worth the hassle to build a cluster.
|