Osztott 2D raszterizációs modell többmagos processzorok számára

Abstract

The graphics processing unit (GPU) has become part of our everyday through desktop computers and portable devices. Because of the dedicated hardware the visualization has been significantly accelerated and today's software uses only the GPU for rasterization. Beside the graphical devices, the central processing unit (CPU) has also made remarkable progress over. The multi-core architectures and new instruction sets have been appeared. This paper aims to investigate how effectively this multi-core architecture can be applied at the two-dimensional rasterization process, what are the benefits and bottlenecks of this rendering model. We answer the question, that would it be possible to design a software rendering engine to meet the requirements of today's computer games.