Colloquium Series

Fall 2025

December 5, 2025

Sunil Kumar, Tufts University

Topic: Re-Solving Heuristics for Network Optimization Problems
Time: 4:00-5:00 pm
Location: Barnum Dana Complex, Room 104
Reception: JCC501
Abstract: We consider a class of dynamic network optimization problems known as revenue management problems. These arise in many applications including airline transportation. We study the performance of a class of heuristic control policies that solve a much simpler static optimization problem periodically. We bound the loss in performance of these policies from the optimal policy and show that this bound remains constant even as the problem size grows.  

November 14, 2025

David Keyes, KAUST

Topic: For What Should the Bell Toll?
Time: 4:00-5:00 pm
Location: JCC265
Reception: JCC501
Abstract: Over its 37-year history, the Gordon Bell Prize has strongly influenced the development of high-performance simulation and big data statistics in mostly positive but also in some negative ways, with implications for the operations of leading supercomputer centers. We briefly discuss the evolution of the prize in its several awarded categories, then focus on a handful of submissions of influence. For example, the 1999 Gordon Bell Special Prize paper that first addressed an unstructured grid PDE application is presented as a precursor of the 2009 roofline model now universally used to reveal whether an application is compute- or bandwidth-bound, or even instruction issue-bound at a finer examination. The flop/s rate orientation of the Prize has encouraged wasteful flops, where a smaller number would provide more cost-effective delivery of quantities of scientific interest to working accuracy. In support of the latter, we provide examples from four of our own recent submissions that were selected as finalists, and one that took the 2024 Gordon Bell Prize in Climate Modeling. How to pilot the Prize at the confluence of the roaring rivers of simulation and learning, with the tributary of quantum computing about to join, is the final topic of discussion. As energy costs now dominate what can be simulated and what can be learned or inferred, the admittedly elusive metric of “science per Joule” is proposed as one of several aspects to be addressed in the future evolution of this inspiring legacy.

October 17, 2025

Margaret Beck, Boston University

Topic: Topological Implications for Stability
Time: 4:00-5:00 pm
Location: JCC265
Reception: JCC501
Abstract: In the context of many PDEs, the stability of a solution roughly corresponds to whether or not that solution can be expected to be observed in the long-time dynamics of the system. Determining stability often amounts to understanding the point spectrum, or eigenvalues, of an associated linear differential operator, with positive-real-part eigenvalues indicating an instability. In this talk, we will discuss recent work connecting the topological properties that govern the existence of the solution with the existence of positive eigenvalues and instability. 

September 5, 2025

John Urschel, MIT

Topic: Nodal Statistics for Graphs and Matrices
Time: 4:00-5:00 pm
Location: JCC265
Reception: JCC501
Abstract: The study of discrete nodal statistics, that is, data regarding the zeros of Laplacian eigenvectors, provides insight into structural properties of graphs and matrices, and draws strong parallels with classical results in analysis for Laplacian eigenfunctions. In this talk, we will give an overview of the field, covering key results on nodal domains and nodal counts for graphs and their connection to known results and open problems in the continuous setting. In addition, we will discuss some recent progress towards a more complete understanding of the extremal properties of the nodal statistics of a matrix.