Institut d'Électronique et de Télécommunications de Rennes

| ENT |

Accueil du site > Accueil > Production scientifique > Thèses en cours > Équipe VAADER (Dpt IMAGE)

Energy Optimization for Video Processing on Embedded MPSoCs

Doctorant : Erwan NOGUES.

Directeur de thèse : Daniel MENARD.

Thèse débutée le : 01/11/2013.

The proposed PhD thesis will require an in-depth study of power estimation and optimization, dataflow programming and distributed signal processing systems. The first objective of the PhD thesis will be the integration of energy consumption estimators into rapid prototyping tools developed in the laboratory5. Estimation of energy consumption will be based on hybrid approaches combining profiling information and measurements6. These estimators will then be used to optimize the system energy consumption via methods such as Dynamic Voltage and Frequency Scaling (DVFS) and power gating. If these techniques prove efficient, hardware/software co-design methods will be applied to optimize the system conception by considering energy consumption early in the system design process. Finally, energy optimization at algorithm level will be investigated. Applications parameters and algorithms will be tuned to reduce the energy consumption.

The main target platform will be the Keystone Digital Signal Processors from Texas Instruments. The main target application will be MPEG HEVC video compression. The student will be involved in collaborative projects and will benefit from the close relationships between IETR/INSA and the company Texas Instruments.

Page du doctorant

Départements et équipes de recherche