About Genome

Early Genome was a compute farm in the Beowulf tradition; i.e., built with commodity hardware and freely available software. Growing to include more than 60 CPU's, Genome served both as a testbed for CGRB developers and system administrators and as an increasingly powerful tool for OSU faculty, staff and students. Learning from setbacks and capitalizing on successes, Genome has steadily grown to include more than 330 processors. Only about 10% of Genome is currently based on the original commodity hardware. Space and BTU constraints motivated the migration to rack-mount multi-processor nodes and the need to address large amounts of RAM instigated the move to 64-bit hardware. Currently, about 90% of Genome is composed of rack-mount multi-processor nodes that contain 2, 4 or 8 AMD single and dual core Opteron and Intel Xeon quad core based processors.

Two themes have persisted throughout the life of Genome: keeping costs reasonable and facilitating biological computing for the average biologist.

• To mitigate hardware expenses, both to the CGRB and to individual groups of researchers that invest in Genome, the private network for the cluster is based entirely on 1Gb Ethernet. Higher throughput, lower latency network technologies may replace Ethernet in computer clusters, but the cost of hardware components to support Myrinet or InfiniBand can be prohibitive. 1Gb Ethernet has offered a reasonable compromise between network throughput and latency, and hardware cost.
• Most Genome users fit into one of two categories: those who know Genome exists and those who don't. Several individual laboratories at OSU have built research programs that rely on high-performance computing; for example, The Arabidopsis Small RNA Project and the Marine Microbial Genomics project. These groups have invested in hardware and developed custom algorithms and software to capitalize on Genome's capabilities. However, the majority of Genome users may never realize they are using the cluster. The CGRB maintains an extensive suite of web-accessible bioinformatics software. To accommodate the number of potential bioinformatics users, most of the jobs are submitted to Genome. The CGRB Bioinformatics website is one of the easiest ways OSU biologists can take advantage of Genome to advance their research.