RDISK: a reconfigurable architecture for browsing genomic libraries

For several years, the continued progress of biotechnology has amassed a considerable body of genomic data, thus providing the ability to perform whole-genome sequencing much faster. RDISK architecture is a prototype system consisting of a cluster of forty-eight specialized nodes, each comprising a hard disk closely connected to a field-programmable gate array (FPGA). The aim is to filter these genomic databases "on-the-fly", that is to say, without slowing down the reception and reading of information coming from the disks. Depending on the nature of the queries, the system is reconfigured automatically.