One nice thing about using a pipeline is that it will be multi-threaded by default. That is each process will have it's own thread(s) and distribute any work a

One nice thing about using a pipeline is that it will be multi-threaded by default. That is each process will have it's own thread(s) and distribute any work across multiple cores. Also with something like Gnu Parallel you can easily scale across multiple machines. Maybe there are more efficient approaches but a pipeline and gnu parallel is pretty easy to scale up without thinking much about it.