PhD Position in Computational Storage and Scientific Computing

This PhD position is hosted by the KerData research team at the Inria center in Rennes, France, in collaboration with international partners including Otto von Guericke University Magdeburg. The research focuses on investigating mechanisms to formalize, capture, model, and evaluate computational storage in distributed environments to address data movement bottlenecks in scientific computing and AI. The candidate will perform state-of-the-art surveys, establish testbeds using simulation and hardware emulation, and develop methodologies for large-scale performance modeling. The project involves close cooperation with domain scientists and potential international research visits.

• Conduct a comprehensive literature survey and define ontologies and taxonomies for distributed computational storage systems across application, middleware, and system perspectives.
• Establish technical foundations to empirically assess hardware performance and analyze workflow task characteristics.
• Implement and maintain computational storage testbeds utilizing simulation, hardware emulation, and physical hardware stacks.
• Develop a methodology to faithfully model performance extrapolations to large-scale deployment scenarios.
• Collaborate with domain scientists to analyze real-world scientific computing use cases and identify computational storage opportunities.
• Contribute to software prototypes and validate approaches using real-world applications on platforms like Grid5000 or Chameleon Cloud.