Multi-core Architectures and Programming
Multi-core processors offer a high theoretical computation performance and therefore open up new fascinating possibilities in scientific and other domains, like multimedia, medicine and finance. In order to fully exploit the performance, an efficient mapping of algorithms to the architecture of the respective multi-core processor has to be determined. Compared to traditional single-core processors, a radical rethinking of programming methodologies needs to be undertaken. Our aims are to gain new insights into modern multi-core architectures and the corresponding programming paradigms in order to exploit the platform and to make it easier to map algorithms to the platform. As platforms, we focus in particular on graphics cards from NVIDIA (e.g. Tesla based systems), on systems based on the Cell Processor (e.g. Sony’s PLAYSTATION 3). However, we look also for new architectures like Tilera’s TILEPro64 and Intel’s Many Integrated Core (MIC).
Mashup of different Multi- and Many Core Architectures
No publications found.