The program consists of five modules, and each module balances conceptual learning with extensive hands-on practice. Each day includes two or three lectures and hands-on training, i.e., lab or reading and discussion. The labs reinforce the theoretical material through practical exercises using modern scientific computing frameworks.
Students are expected to develop a deeper understanding of mathematical concepts, solver techniques, convergence theory, and practice implementation skills. The activities include:
(i) run simulations in Firedrake, a system that allows experimenting with large-scale (nonlinear) PDE problems without extensive programming;
(ii) work with HPC tools and GPU accelerated workflows;
(iii) analyze solver performance and convergence behavior; and
(iv) integral project: use the learned methods to solve a real problem.
Monday (06/14):
Tuesday (06/15):
Wednesday (06/16):
Thursday (06/17):
Friday (06/18):
The second week shifts to advanced preconditioning methods, multiphysics coupling strategies, and practical applications. Most of the topics are rarely covered in standard university curricula.
Monday (06/21):
Tuesday (06/22):
Wednesday (06/23):
Thursday (06/24):
Friday (06/25):
Module 1:Basic Topics on Nonlinear Problems and Solvers
Module 2: Advanced Nonlinear Solvers
Module 3: Preconditioning Strategies
Module 4: Coupled Multiphysics Systems
Module 5: High-Performance Computing