Job: PhD Positions in Scientific GPU Computing, Denmark

Study Subject: Engineering, Computational Science and Engineering, Applied Mathematics

Employer: Technical University of Denmark

Level: PhD

Scholarship Description: This PhD project focuses on computational methods for optimization of dynamic systems. The optimization algorithms used for that will be based on interior-point methods and convex optimization. The computational bottleneck in these optimization algorithms is often the matrix computations, and we aim to develop GPU-based high-performance algorithms that overcome this bottleneck. One of our test problems is distribution and control of electrical power via a “smart power-grid” through the use of model predictive control; but our developments will also produce advances in other areas of constrained optimization.

The research will be carried out in an interdisciplinary environment in collaboration with one other PhD and one post doc within the research project, and with guidance of the project’s research team.

Requirements
Candidates must have a master degree in engineering, computational science and engineering, applied mathematics, or equivalent academic qualifications. We are interested in candidates with experience in some (not necessarily all) of the following areas

* Numerical simulation, optimization, control
* Programming (Matlab, C/C++, and/or Fortran)
* High-Performance computing
* Multi-core computing
* Scientific computing
* Convex optimization and matrix computations

You will get the necessary training and courses in the areas that you do not master yet.

A good command of the English language is essential.

Scholarship Application Deadline: 15October2010

Further Scholarship Information and Application:
http://www.dtu.dk/English/About_DTU/vacancies.aspx?guid=e4a48f4d-b536-4e53-8015-845db962fca0