Improving system energy consumption through optimizing memory usage

Affiliated to

Location

Timing

Description

An exploration in the area of explicitly managed memory hierarchies. Focusing on how to use memory hierarchy management to affect the energy consumed when running an application. Constructing a predictive model for energy consumption based on empirical measurements for Cortex-R4. Validating the model using application profile data and using it to implement methods, using binary rewriting/link-time optimisation, that
improve total power consumption for an application.

Essential Requirements
PhD student in computer architecture, compilers or related fields. A keen interest in low-level embedded software
Detailed knowledge of software development tool chains for ARM including C/C++ compiler, assembler and linker
Good programming skills in C/C++

Desirable knowledge/skills
Binary rewriting/link-time optimization for ARM ISA
Experience in software development using Java, Python or Perl