Reconfigurable and Parallel Architecture for Browsing Genomic Databases
Context
Scanning the genomic database is a common task, daily performed by thousands of researchers for extracting new knowledge from this huge amount of data. Today, this represents tens of Gigabytes. Tomorrow, according to the genomic data exponential growth, the volume of information to manipulate will reach Terabytes.
RDisk System
We propose a 48 parallel disk system whose originality is to attach reconfigurable computation capabilities near the disk for providing on-the-fly data filtering. Speed-up is provided both by accessing tens of disk in parallel and by hardwiring efficient FPGA filters.
Applications
- BLAST-like search
- Motif search
- Content-based image search
Bibliography
- A. Noumsi, S. Derrien, P. Quinton, Acceleration of a Content Based Image Retrieval Application on the RDISK Cluster. IEEE International Parallel & Distributed Processing Symposium, April 2006 (pdf)
- D. Lavenier, S. Guyétant, S. Derrien, S. Rubini A reconfigurable parallel disk system for filtering genomic banks, ERSA’03, Engineering of Reconfigurable Systems and Algorithms , Las Vegas, Nevada, USA, 2003. (pdf)
- S. Guyetant, M. Giraud, L. L’Hours, S. Derrien, S. Rubini, D. Lavenier, F. Raimbault, Cluster of re-configurable nodes for scanning large genomic banks, Parallel Computing, 31 (1) 2005 (link)
- S. Guyétant, Architecture parallèle reconfigurable pour le filtrag de banques de données non structurées ; application à la génomique Thèse de l’université de Rennes 1, décembre 2004 (link)
- M. Giraud, S. Guyétant, D. Lavenier, Encodage Linéaire d’automates pondérés. Filtrage de motifs génomiques et application sur l’architecture prototype R-disk, TSI,24(6), 2005.
