From a practical point of view, parallel computing appears to be in some kind of crisis: in spite of some earlier optimistic expectations, parallel processing is still not a common matter, and the announced "second computer revolution" that should have been caused by replacing sequential computers with parallel ones is not even in sight. The reasons of this unfortunate state of art have been convincingly identified to lay in theory. Namely, so far computer science did not provide any reasonable abstract model of a parallel computer that would be at the same time good both as a conceptual tool for software development and as a realistic model amenable to existing parallel hardware technologies.