*By Professor David Keyes *

*Director of the Extreme Computing Research Center **(KAUST)
*

The ball is in the court of mathematicians! After a 25-year stasis in programming models, during which computational scientists and engineers enjoyed an exhilarating ride of more than 7 orders of magnitude of hardware performance while simultaneously shedding whole powers of N (the discrete problem size) in arithmetic complexity off of common mathematical tasks, new algorithms to use next-generation computers close to their potential are required. Instead of squeezing out flops, algorithms must now put the squeeze on synchronizations, memory, and data transfers, while extra flops on locally cached data represent only small costs in time and energy. Algorithmic capabilities (to make forays into, e.g., data assimilation, inverse problems, uncertainty quantification, and the merger of analytics with simulation) must be co-designed with the hardware and the new programming models required to use it. We briefly recap the architectural constraints and application opportunities. We then concentrate on two types of tasks, each of consumes a large portion of today’s supercomputing cycles: large dense symmetric/Hermitian linear systems (covariances, Hamiltonians, Hessians, Schur complements) and large sparse Poisson/Helmholtz systems (solids, fluids, electromagnetism, radiation diffusion, gravitation). We examine progress in porting solvers for these tasks to the hybrid distributed-shared programming environment, including the GPU and the MIC architectures that make up the cores of the top scientific computers “on the floor” today in G-20 countries, including Saudi Arabia, and “on the books” for delivery by 2022. Be prepared: austere architectures ahead!

**Biography**: David Keyes is the director of the Extreme Computing Research Center at King Abdullah University of Science and Technology, where he was a founding dean in 2009, and an adjunct professor of applied mathematics at Columbia University. Keyes earned his BSE in Aerospace and Mechanical Engineering from Princeton and his PhD in Applied Mathematics from Harvard. He works at the algorithmic interface between parallel computing and the numerical analysis of partial differential equations. He is a Fellow of SIAM and AMS and has received the AMC Gordon Bell Prize, the IEEE Sidney Fernbach Award, and the SIAM Prize for Distinguished Service to the Profession.

### More Information:

**For more info contact: **Professor David Keyes : email: david.keyes@kaust.edu.sa

Date: Thursday 21^{st} Sep 2017

Time:12:00 PM - 01:00 PM

Location: Building 9, Lecture Hall 1 Room 2322

Refreshments: Light Lunch will be served at 11:45