Accelerator technology for structured mixed-integer programming problems (ESA practice)

Add to your list(s) Download to your calendar using vCal

• Andrea Picciau (Imperial College)
• Monday 23 June 2014, 11:00-12:00
• Level 9 Mahanakorn Lab., EE Dept..

If you have a question about this talk, please contact Grigorios Mingas.

Mixed-integer programming problems (MIP) are a class of optimization problem that is very important for its applications to control engineering. Algorithms for solving this kind of problem are usually very computationally intensive. With computing technology moving towards many-core systems, it is necessary to understand how mathematical programming can make the best use of massively parallel devices, such as GPUs and FPG As. The goal of this research is to show that by capturing the structure and parameters of mixed integer programs in a suitable language, it is possible to automatically generate accelerators with better performance than unaccelerated computers or accelerators that do not account for the structure. In this presentation, some preliminary experiments on MIP example problems coming for the control engineering area are illustrated. After an analysis of their results, a timeline for the development of the project is proposed.

This talk is part of the CAS Talks series.

This talk is included in these lists:

• Andrea Picciau's list
• CAS Talks
• Circuits and Systems Group: Internal Seminars
• Level 9 Mahanakorn Lab., EE Dept.

Note that ex-directory lists are not shown.