I am a Research Scientist in IBM Dublin Research Lab in the High Performance Systems Group. My recent work focused on distributed I/O acceleration and dynamic resource management for the IBM Power based heterogeneous HPC systems, more specifically for the Human Brain Pre-Commercial Procurement project, leading to several patents and external publications. I also contributes to a many-task runtime targeting heterogeneous platforms through the EU H2020 FETHPC-01 Allscale project (www.allscale.eu)
I defended my PhD on fault tolerance for MPI communication libraries under the supervision of Franck Cappello and Joffroy Beauquier in 2006. I worked on fault tolerance protocols for MPI libraries through the MPICH-V and OpenMPI projects and their runtime environments. I also contributed to the DAGuE project on scalability and performance issues of runtime environment for large scale systems under the supervision of Jack Dongarra during my time as a Research Associate in the Distributed Computing group of the Innovative Computing Laboratory.
My areas of interest include Fault tolerance, Scalability, Middleware for large scale HPC systems.