The Open Storage Network will enable researchers to manage big data more efficiently than ever before.
The South Big Data Hub announced its selection as one of four regional big data hubs awarded an $1.8 million grant from the National Science Foundation. The company will use this grant for a two-year project to develop a data storage network.
A collaborative team will use its expertise and resources to develop the Open Storage Network. The network will allow academics nationwide to work together and share their data quickly and easily.
Project Leaders & Participants
Alex Szalay of Johns Hopkins University will help this project with the assistance of data storage partners across the United States including the National Data Service and members representing each of the four NSF-funded Big Data Regional Innovation Hubs (BD Hubs):
- South Big Data Hub at the Renaissance Computing Institute (RENCI)
- Georgia Institute of Technology
- West Big Data Hub at the San Diego Supercomputer Center (SDSC)
- Midwest Big Data Hub at the National Center for Supercomputer Applications (NCSA)
- Northeast Big Data Hub at the Massachusetts Green High Performance Computing Center (MGHPCC)
- Pittsburgh Supercomputing Center (PSC)
See also: Managed services looks to plug big data skills gap
“There are hundreds of issues to solve before we have a “datanet” as efficient and cooperatively well organized as the internet,” Christine Kirkpatrick, executive director of the National Data Service and co-chair of the Big Data Hubs’ Data Sharing and Cyberinfrastructure Working Group said. “But just as there are many complexities to solve in the long-term, these solutions can be facilitated by the simplest of additions – here the establishment of a connected storage network. Sharing, reproducibility, replication, and data management all fundamentally rely on a place to store data.”
The data transfer systems for the new data storage network are designed to match the speed of a 100-gigabit connection while being low-cost, large capacity and high bandwidth while utilizing a small number of nodes.
