Meeting Times:  Wednesdays, 12:05-12:55 pm
Location:  CCB 102

Instructor:  Edmond Chow
E-mail:  (the best way to reach me)
Office Hours:  By appointment (or drop by my office)

This is a graduate seminar/reading course for 1 credit and will meet for 1 hour per week. The course is already fully subscribed, but you can come to the first class if you wish to register.



Course Description

Solvers are the mathematical workhorses at the core of any numerical simulation or scientific code. They are often general and applicable to problems from many different applications, but they are also sometimes specialized to specific physical problems. Efficiency, robustness, and accuracy of solvers are of paramount importance.

Topics may include:

  • Direct solvers for sparse linear systems
  • Preconditioning for iterative linear solvers
  • Multigrid, including algebraic multigrid and nonlinear multigrid
  • Large-scale eigensolvers (for selected eigenvalues and eigenvectors)
  • Solvers for constrained optimization

Other possible topics include: methods for solving differential-algebraic equations and efficient algorithms for computing long-range interactions in particle-based simulations.

Each week, we will read a classic paper or other pedagogical text and the instructor will present and lead a discussion of its basic ideas.


Expectations and Grading

You are expected to read approximately 12 papers or texts this semester, and to get as much out of them as you can! Your grade will be based on your participation in class discussions.


Course Materials

Contact

Edmond Chow
Georgia Tech College of Computing
KACB 1332
(404) 894-3086