Current technology provides a four orders of magnitude difference between the energy cost of executing one instruction and that of sending one packet over the air. Future networks of high density will therefore require radically different paradigms of communication and computation. The challenges they bring stem from a twofold scalability problem. The first one is that of the network scalability: the sheer number of nodes makes it difficult to use traditional methods with a human is involved in the management loop. The second problem is with the scale of the node: as nodes become increasingly smaller, their power capacities are reduced, so computation and communications need to be adjusted as such. In this work we look at position as a basic building block in designing networks of high density. First, for nodes which may not use GPS provided positions, we describe methods to derive either absolute positions or mappings to Euclidean space that can be used as relative positions. Second, we describe services that can use node positions to better design and manage large dense networks. These approaches favor computation over communication to reflect these new technological trends.