I am a Full Professor of Applied Mathematics at Université Grenoble Alpes, affiliated with the Laboratoire Jean Kuntzmann and the AIRSEA research team. My work lies at the intersection of numerical analysis and high-performance computing (HPC), with a focus on the efficient simulation of geophysical flows relevant to atmospheric and oceanic dynamics.
My research centers on the design and analysis of advanced numerical methods for partial differential equations, particularly parallel-in-time integration techniques, exponential integrators, and high-order time-stepping schemes. I apply these methods to challenging large-scale problems such as the shallow-water equations on rotating spheres, which are fundamental in climate and weather modeling. A key contribution of my work is improving the stability and efficiency of solvers for stiff and oscillatory systems, enabling accurate simulations over long time horizons. A major theme of this research is the extension of parallelism beyond spatial decomposition. I have contributed to the development and analysis of parallel-in-time algorithms, including methods such as Parareal, multilevel approaches, and rational approximations, which allow the temporal dimension of simulations to be exploited for concurrency. These advances address critical scalability limitations in modern HPC systems and open new perspectives for large-scale scientific computing.
In parallel, I have worked extensively on the co-design of numerical algorithms and computing architectures. My contributions include research on dynamic resource management, malleability/elasticity in HPC systems, and asynchronous many-task runtime environments. By integrating algorithmic innovation with system-level considerations, my work enables efficient execution on distributed and heterogeneous platforms, including CPU- and GPU-based supercomputers.
My publication record reflects these core themes, with contributions spanning exponential integration methods, rational approximation techniques, parallel-in-time schemes, and HPC runtime systems. My work also extends to applications such as seismic wave propagation, adaptive grid-based simulations, and broader geophysical fluid dynamics, demonstrating a strong link between theoretical development and real-world scientific challenges.
Beyond research, I am actively involved in the international scientific community. I have served on program committees for major conferences in computational science and high-performance computing, including events such as Supercomputing, the International Conference on Computational Science, and Euro-Par. I have also co-organized workshops and initiatives on emerging topics such as malleability, resource management in HPC, and numerical methods, fostering collaboration between mathematicians, computer scientists, and domain scientists.
In addition to my research and service activities, I am committed to teaching and academic mentoring in applied mathematics and scientific computing. My teaching covers numerical methods for PDEs, time integration techniques, and HPC, contributing to interdisciplinary training at the interface of mathematics and computer science.
Overall, my work advances the development of scalable numerical methods and computing techniques for next-generation scientific simulations, with a particular impact on climate and geophysical modeling.
If you need anything or want to kick off a collaboration, don't hesitate to contact me.
Martin
