Degree Name

Master of Engineering - Research


School of Electrical, Computer and Telecommunications Engineering


Real-time graphics processing on the cloud poses significant challenges in terms of processing capability, data transmission, and the management of latency. The rendering of large and complex graphics data requires large processing power and significant storage that low-powered machines are unlikely to handle capably. In addition, the transmission of graphics may introduce considerable delays, leading to poor interactivity. Numerous works have been carried out taking these issues into account, most of which being based on level of detail (LOD) and image based rendering (IBR) techniques. However, there are many tradeoffs that need to be carefully studied in order to realize some of the benefits of cloud computing for three dimensional (3D) networked graphics. In this project, we explore the state of the art remote rendering, or in other words, moving the rendering of complex graphics data into a cloud system. A networked rendering paradigm based on our proposed pipeline-splitting method is introduced to facilitate a remote-rendering system with the aim of partitioning the rendering workload between the client and server. We also propose a visibility streaming method for networked applications to reduce the transmission capacity required. One of the main advantages of our proposed methods is that it is easy to scale up at the server side by distributing the workload to be handled in different machines, leading to a significant improvement at the server side in terms of performance.