The CRS4 DC group's main research interest is developing new computing solutions for heavy duty, data-driven scientific applications. We are interested both in how one should cast the problem at hand so that it will better fit a given computational infrastructure, and in how one can adapt a given computational infrastructure to efficiently solve it.

Representative examples of current active research domains are the following:

• SOA based control planes for the dynamic allocation of computational infrastructures, with the goal of providing a control plane architecture that can handle the allocation and configuration of network and computational resources with a uniform approach;

• virtual clusters for HPC applications. More specifically, we are interested in the use of virtual clusters for large scale data-driven applications, where these flexible tools promise to provide an elegant mechanism to move computation where data reside;

• specialized parallel computing frameworks, map-reduce as an example, designed to handle large scientific data-driven applications with particular emphasis on biomedical data.

Our research activities harness the CYBERSAR and CRS4 computational infrastructures both as experimental and deployment platforms.

Our work is partially supported by the Italian Ministry of Research under the CYBERSAR project.