A flexible, scalable approach to real-time graphics

Shrubsole, P.A., 2000. A flexible, scalable approach to real-time graphics. PhD, Nottingham Trent University.

[img]
Preview
Text
10183025.pdf - Published version

Download (22MB) | Preview

Abstract

There have been several implementations of multiprocessor graphics systems in recent years that achieve high levels of realism at high performance. However, it is common for these implementations to be deficient either in their flexibility, so that they become application dependent, or in their scalability, resulting in a limited shelf life.

The aim of this research project is to design and investigate a graphics architecture that is flexible and scalable, whilst providing high quality output in real-time.

To achieve this, the thesis first focuses on high performance antialiasing techniques. New algorithms for performing texture antialiasing called texture potential mapping are presented which provide very high quality output at high performance and can easily be incorporated into multiprocessing environments.

The thesis then focuses on multi-processing architectures that can incorporate these advanced rendering algorithms in a flexible way whilst achieving scalable real-time performance. The latter part of the thesis presents a new architecture called the cellular array that embodies both flexibility and scalability in its design. Simulation experiments of the cellular array are presented that examine the behavior of this architecture for different applications and provides strategies for ensuring that the multi-processing environment is consistently well balanced.

Item Type: Thesis
Creators: Shrubsole, P.A.
Date: 2000
ISBN: 9781369313178
Identifiers:
NumberType
PQ10183025Other
Divisions: Schools > School of Science and Technology
Record created by: Linda Sullivan
Date Added: 28 Aug 2020 12:29
Last Modified: 21 Jun 2023 08:04
URI: https://irep.ntu.ac.uk/id/eprint/40575

Actions (login required)

Edit View Edit View

Views

Views per month over past year

Downloads

Downloads per month over past year