What piece of technology will we have in our hand in five years? The mobile phone has become a standard personal electronic item. A likely scenario is that the mobile phone will evolve into a personal multimedia platform in a not so far future. In order to incorporate current multimedia implementations on the limited resources available in mobile phones today, these implementations must be evaluated for specific hardware and software requirements.
This thesis investigates the incorporation of the interactive 3D-graphics media on small platforms with limited resources, like a mobile phone. 3D-graphics typically make high demands on resources such as memory and processing speed. For PCs, 3D-graphics is often entirely implemented in hardware. For small platforms, limited in memory, processing speed and hardware components, it might be necessary to focus on a split hardware/software solution. The approach presented in this thesis takes a pure software implementation of a 3D-graphics system, measures its performance on a small platform and identifies possible hardware/software partitioning.
Performance is measured by an empirical evaluation and analysis of a spectrum of 3D-application running on the 3D-graphics system. The 3D-applications are written for the OpenGL graphics library and executed with the Mesa graphics library that is very similar to that of OpenGL. They are tested on both an ordinary PC and on an ARM processor, which is a very common processor for small platforms.
Author:
Tobias Ericsson
Supervisors:
Anders Wesslén, ECS
Fredrik Dahlgren, ECS
Mathias Haage, Inst. för Datavetenskap
Presented: 2001-11-14