Power management has been one of the main constraints faced by the designers of all kinds of computer systems in the past decade. Especially in connection to real-time systems its application becomes nontrivial. Due to the vastness of applicable domain, the whole spectrum, from server farms to tiny embedded devices, is poised to benefit from the advances in this field. Over the years it has attracted a lot of interest from research community as well as from industrial players and a considerable body of knowledge has been generated. However, there still remains a multitude of nontrivial challenges, ensuring a high relevance in the years to come.
Analysis of power management techniques require a broad range of specialties i.e. algorithms, network calculus, microelectronics, scheduling theory, financial analysis to name a few. In this seminar we will survey the existing body of research in this domain and identify the important scientific and economic challenges. Areas covered will be fairly broad and will equip participants with necessary knowledge to analyze the novel approaches in the field.
A maximum number of 10 participants are allowed. Each of them will be expected to go through the relevant literature and prepare a scientific report of about 15 pages on the assigned topic and present the findings through a presentation of 20 minutes in the end of the course. All the presentations will be held at the end of the semester in block form. The topics will be allocated in the first meeting. Most of the administrative issues will be addressed in the first meeting.
Most of the administrative issues will be addressed in the first meeting on 23.10.2013 at 14:00 in room 001, Building 40.28.