The goal of the TADaaM project is to design and build a stateful system-wide service layer for data management in HPC systems. This layer will be twofold. First, it will abstract low level features of thesystem (e.g. topology, network, resource usage) and the software stack (e.g. threads, data). Second, applications will be able to register their needs and behavior thanks to a carefully designed API. With these two sets of information, the layer will optimize the execution of all the running applications in a coordinate fashion and at system-scale.
The TADaaM team was created after the end of the Runtime Inria team-project.