Geoffroy Vallee

Dr. Vallee is R&D Staff Member in the Computer Science Group of the Computer Science and Mathematics Division at the Oak Ridge National Laboratory (ORNL), Dr. Vallee is focusing on operating systems, system resilience and system run-time for high-performance computing (HPC) systems, focusing on cluster and grid computing, system-level virtualization, tools, and High Availability (HA) issues.
Dr. Vallee is working on various projects in the context of HPC, including research in run-time systems for MPI application in the context of a co-design effort (Institute for Advanced Architectures and Algorithms) and the development of a scalable run-time system (Scalable Tool Communication Infrastructure). Furthermore, Vallee's work on system tools and OS led to the development of a fully integrated and easy to install software bundle designed for high performance cluster computing, which are fully integrated into the OSCAR cluster management suite.
For instance, SSI-OSCAR provides is based on Vallee's expertise in SSI systems from his Ph.D. thesis contributions (Vallee is a co-designer of the Kerrighed SSI). In the context of these efforts, he has integrated the OSCAR team and is now chairman of the project. His Ph.D. research has on distributed operating system, focusing on system-level mechanisms and policies for the efficient management of processes and threads in the context of a SSI. This effort led to the design and implementation of a adaptable global scheduler for clusters. This work has been leveraged for the creation of a industrial product, nowadays developed and maintained by the KerLabs company (http://www.kerlabs.com/).