This paper assesses the empirical usefulness of some recent developments in the theory of spatial and product competition, drawing on the history of product competition in microprocessors over the period 1971-81. In particular, a role is found for agglomeration economies in explaining the observed clustering of designs.