Fine-Grained Parallelism for Genomic Computation
Abstract
Biotechnology progresses lead to an exponential growth of genomic data. This mass of information, available from public data banks, is daily scrutinized by thousands of biologists to extract new knowledge. Powerful computer systems are thus needed to process this huge volume of data. Hopefully, intensive genomic computation is well suited to be implemented on large parallel systems like clusters or grids of computers. This is the solution which is usually adopted by the bioinformatics servers. However, genomic computation performances can be highly improved using a fine-grained parallelization approach through dedicated hardware accelerators. The talk will give an overview of various parallel implementations ranging from VLSI custom chips to reconfigurable FPGA systems.