Modern smartphones are equipped with multi core CPUs and GPUs. They are  capable of running tasks requiring high computing, such as multimedia playback, encoding. For providing satisfactory performance the mobile devices use the maximum clock rates with a single core, or use multiple cores, or use both approaches at the same time. Nevertheless the resulting performance comes at price of high energy cost which is critical for smartphones as they have limited battery capacity. In order to deal with energy inefficiency, ARM proposed big.LITTLE architecture, in which there the “big” core is assigned for high computing tasks such as 3D gaming and the “LITTLE” core is assigned for handing basic telephone-related functions [1].

The purpose of this thesis is augment GPU with the CPU for smartphones. For example, when the smartphone is sleeping, the GPU should do all the background work. In improvement should be done in the Selfish or Android OS kernel. This work requires expertise in Linux kernel development and a dedicated candidate. The development work should be done within 4-6 months. If you have expertise in Linux kernel development, please contact with me.

[1] Big.LITTLE Processing with ARM CortexTM-A15 & Cortex-A7 (Improving Energy Efficiency in High-Performance Mobile Platforms)