Can these parallel programming models be applied at the node abstraction level besides the core abstraction level? By node abstraction level vs core abstraction level, I mean using multiple machines to perform a parallel computation as opposed to using just multiple cores.

Which parallel programming model is most often used in most computer architectures?