Cell Processor Based Sequence Alignment

The project focuses on using the capabilities of Cell processor for computing sequence alignments. In order to achieve this, existing algorithms used for sequence alignment need to be rewritten so that they could be efficiently executed on Cell processor. This process involves making use of two main features of Cell processors - parallelisation and vectorisation, which ensure achieving maximum performance. Two main algorithms for sequence alignment to be analyzed in this project are Needleman-Wunsch (for global sequence alignment) and Smith-Waterman (for local sequence alignment) algorithm. These algorithms are well established on regular PC platforms, but are still not widely used on Cell processors. This project will take full power of Cell processors, and put it in function of these algorithms, analyzing performance improvements and speedup.

Autorentext

Milos R. Boskovic, obtained his Master degree in Network &e-Business Centered Computing (University of Reading, UK) andBachelor degree in Computer Science (University of Nis, Serbia). Has particular interest in Cell processors utilization and ITsecurity.