Congratulations to the following three members of the SIAM community who will receive awards at the virtual SIAM Conference on Applied Linear Algebra (LA21). Additional information about each recipient, including Q&As, can be found below.
Theo Mary
Theo Mary of Computer Lab of Paris 6 (LIP6) is the 2021 recipient of the SIAM Activity Group on Linear Algebra Early Career Prize. He will give a presentation at the SIAM Conference on Applied Linear Algebra (LA21), to be held in a virtual format May 17 – 21, 2021.
The prize is awarded to Mary for his significant contributions to linear algebra topics including block low rank methods, software development, probabilistic rounding error analysis, mixed precision arithmetic, and backward error analysis.
The SIAM Activity Group on Linear Algebra (SIAG/LA) awards this prize every three years to one post-Ph.D. early career researcher in the field of applicable linear algebra for outstanding contributions to the field within six years of receiving the Ph.D. or equivalent degree as of January 1 of the award year.
Theo Mary is a CNRS researcher at the LIP6 laboratory of Sorbonne University in Paris, France, since 2019. He earned his Ph.D. degree from the University of Toulouse in 2017. His Ph.D. thesis, which he prepared at the IRIT laboratory under the supervision of Patrick Amestoy and Alfredo Buttari, was awarded an honorable mention for the Householder prize. He then spent two years at The University of Manchester, where he worked as a postdoctoral researcher in Nick Higham's numerical linear algebra group.
Q: Why are you excited to win the SIAG/LA Early Career Prize?
A: I am deeply honored to receive this prize and I am grateful to SIAM for sponsoring it. I'm really excited that the research that my colleagues and I have been working on is recognized by the community.
Q: Could you tell us a bit about the work that won you the prize?
A: My research focuses on the development of high-performance numerical algorithms, and on their analysis in finite precision arithmetic. I have particularly worked on accelerating the solution of linear systems on modern parallel supercomputers, by means of two approaches: block low-rank approximations and mixed precision arithmetic. Block low-rank methods exploit the special structure of data sparse matrices, which arise in a wide range of applications, to obtain solution algorithms with nearly linear complexity. Such methods can thus tackle very large-scale problems that would be intractable with standard approaches. Mixed precision methods combine standard double precision arithmetic with the lower precisions that are newly available on modern hardware, especially half precision (16-bit arithmetic). Such methods can achieve tremendous performance gains without sacrificing the accuracy and stability of the computations. Most recently, we have been looking at combining both approaches, low rank approximations and mixed precision arithmetic, with very exciting early results.
Q: What does your work mean to the public?
A: One of the aspects that I find most exciting of working in numerical linear algebra is its incredibly wide range of applications. Linear systems are ubiquitous in scientific computing, and I have had the opportunity to apply my research in several exciting fields and to work with great people from the application domains and industry: in geosciences, electromagnetics, structural mechanics, to name a few. To the general public, this probably does not mean much, but, whenever they drive their car, take the airplane, or rely on electricity, somewhere along the chain of production, linear systems were solved!
Q: What does being a SIAM member mean to you?
A: I really enjoy being part of SIAM. The SIAM journals are at the top of their respective fields, and my experience with them, both as an author and a referee, has been consistently smooth and pleasant. And the SIAM conferences are always great opportunities to meet the rest of the community and learn about new exciting advances.
Michel Crouzeix and César Palencia de Lara
Michel Crouzeix of Université de Rennes and César Palencia de Lara of Universidad de Valladolid are the recipients of the 2021 SIAM Activity Group on Linear Algebra Best Paper Prize. They will present their paper at the SIAM Conference on Linear Algebra (LA21) to be held in a virtual format May 17 – 21, 2021. The prize is awarded to the pair for their paper "The numerical range is a (1 +√2)-spectral set''.
The SIAM Activity Group on Linear Algebra (SIAG/LA) awards this prize awarded every three years to the author(s) of the most outstanding paper, as determined by the selection committee, on a topic in applicable linear algebra in the four calendar years preceding the award year.
Michel Crouzeix is Emeritus Professor. He received his Thèse d’état at Paris 6 University and was Professor at Rennes 1 University for 30 years. His research interests were mainly in Numerical Analysis: finite element methods, discretization of PDE and ODE, bifurcation theory, Davidson’s method, shape optimisation. Stability problems in time discretization has led him to work in functional calculus.
César Palencia de Lara is Professor at the University of Valladolid (Spain). His research interest is focused on Numerical Analysis and its interplay with Functional Analysis. Together with stability and convergence issues for time discretizations of PDEs and evolutionary equations with memory, he has also considered ill-posed problems, the numerical inversion of the Laplace and some other problems in image processing.
Q: Why are you both excited to receive the SIAG/LA Best Paper Prize?
A: We are very proud that our paper has been chosen for this Prize and we wish to express our deepest gratitude to SIAG/LA for honoring us with this recognition. It is very rewarding for us to see that our result is useful for people working in Applied Linear Algebra and, even though it is not yet optimal, we hope that it contributes to renew the interest in the subject.
Q: Could you both tell us a bit about the research that won you the prize?
A: In a seminal article, B&F Delyon have obtained a generalization of the fascinating von Neumann inequality. Their paper inspired an addictive conjecture, formulated by M. Crouzeix, that has generated a number of researches during the last seventeen years. Though the conjecture has not yet been solved, our paper means a substantial improvement of the previous results and hopefully it provides a better understanding of the question.
Q: What does your work mean to the public?
A: We are looking for an estimate that can be expressed in a very simple way but that is particularly useful for the analysis of stability in a non-self-adjoint context. It has raised the interest of people working in a variety of fields, such as linear algebra, numerical analysis, probability, operator algebra, harmonic analysis, group and semigroup theory, and more. We hope that our contribution would render the theory accessible to a larger public.
Q: What does being a member of SIAM mean to you?
A: For us, as for so many other people, SIAM is nothing but the most prestigious and successful worldwide society in applied mathematics. SIAM is the model upon which other societies (in particular, SIAM in France and SEMA in Spain) have been founded. We appreciate the high level and interest of the thematic SIAM Journals and the way SIAM drives the collaboration in the community of applied mathematicians, by means of its activity groups, the other publications, the congresses. Such a liveness is very encouraging and give us the feeling that together we are stronger.