Case Study: University of Southern California achieves petabyte-scale, gigabyte-per-second genomics throughput with DataDirect Networks GRIDScaler

University Uses a High Performance, Scalable Infrastructure To Support Next-Gen Genomics Sequencing

The Zilkha Neurogenetic Institute at USC, led by Dr. James Knowles, needed to scale genomic research but was constrained by a remote legacy SAN and NFS bottlenecks that limited transfers to 30–50 MB/s. With three Illumina HiSeq2000 sequencers ramping to produce terabytes per day and limited IT headcount to deploy new systems, the lab required a high‑performance, single‑namespace solution capable of >1 GB/s throughput and petabyte scalability to avoid slowing time to discovery.

USC and DDN deployed a GRIDScaler appliance (SFA10K‑E) with a parallel file system, 10GbE connectivity and a local caching gateway, expanding raw capacity to ~1.2 PB (800 TB usable) and supporting both parallel storage and NFS. The new system lets the lab run Illumina CASAVA and multiple BWA alignment workflows concurrently while uploading data, cutting processing/upload time to about one‑third, simplifying management for HPCC staff and providing a clear path to multi‑petabyte growth.

University of Southern California

James Knowles

Bioinformatics Programmer and Analyst, Department of Psychiatry & the Behavioral Sciences