The concentration of resources is a core subject of social science and structures most of the current political debates. Recent studies in economics and sociology have improved knowledge on such concentrations by using exhaustive population via register data. The use of traditional units of analysis (individuals, companies, parent-child relations, households), however does not fully take advantage of the possibilities offered by these data. This project offers a unique and significant renewal in the study of resource concentration by combining a traditional idea from anthropology (the structural apprehension of kinship networks) with new tools from mathematics and computer science (network analysis on large graphs via GPU calculation). Family plays a crucial role in resource concentration: wealth is transmitted and inherited within it, and is gathered or disbursed through its union and dissolution. But family also influences non-economic resources that, in turn, influence skills, tastes and aspirations. Our approach defines family as a group of individuals related by parent-child links. Under this definition, the largest family in Swedish registers connects several million individuals. Such metrics open countless opportunities to increase our knowledge on resource concentration, by allowing for the precise distance measurement between every pair of individuals in Sweden. This project will develop this tool and test it on three cases: wealth, occupation and education. The implementation of this research implies the calculation of the shortest distance between each pair of individuals. Such calculation necessitates large computational and storage resources that cannot be managed by traditional computers.