Achieving research progress depends on dealing with increasing volumes of data for pre-processing, analyses, visualization and numerical modeling. Research infrastructures must accommodate those requirements transparently, using heterogeneous backends, provided by e-infrastructure services such as Clouds. Developers need to reach across infrastructure boundaries transparently, while effectively moving data processing and analysis, but not control, away from users and closer to the data. They should deliver transparent combination and optimization of workflows over heterogeneous contexts, such as pipelining HPC codes, data transformation and analytics. Innovative approaches are needed to support the scientific data life cycle in the context of large-scale data and diverse infrastructures.

Presentations are encouraged on, but are not limited to:

• Novel research architectures;
• Workflows and intelligent optimization;
• HPC integration and data-driven approaches, including HPC on the Cloud;
• Provenance;
• Front-ends and services designed for large-scale data-oriented solutions.