In lecture, the justification for bounding speedup to 2 was taking the limit as p goes to infinity. However, in practice if there are only N units of work, we can't split that amongst more than N processors (giving a processor a fraction of a unit of work doesnt make sense). Why do we use the idea of infinite processors for talking about the bounds? Are we just doing something similar to the style of big O, where we only care about orders of magnitude?
In lecture, the justification for bounding speedup to 2 was taking the limit as p goes to infinity. However, in practice if there are only N units of work, we can't split that amongst more than N processors (giving a processor a fraction of a unit of work doesnt make sense). Why do we use the idea of infinite processors for talking about the bounds? Are we just doing something similar to the style of big O, where we only care about orders of magnitude?