Abstract
New high-throughput scientific instruments, telescopes, satellites, accelerators, supercomputers, sensor networks, and running simulations are generating massive amounts of data. In order to be able to exploit these huge volumes of data, a new type of e-infrastructure, the Global Research Data Infrastructure (GRDI), must be developed for harnessing the accumulating data and knowledge produced by the communities of research. This paper identifies the main challenges faced by the future GRDIs, defines a conceptual framework for GRDIs based on the ecosystem metaphor, describes a core set of functionality that these GRDIs must provide, and gives a set of recommendations for building the future GRDIs.