Researchers today work at the computer as much as in the field or in the lab, often grabbing computing resources and applications from research centers and cloud sites across the country or the world. However, data sets are typically centralized, and computation is built around the data as permanent function. As a result, research infrastructure is very expensive to maintain, and hard to share and adapt, even in the age of big data, where research data sets can be widely distributed.
RENCI’s Networking Research Group aims to make distributed, data-intensive research more flexible and collaboration friendly. Through a project funded by the National Science Foundation’s Global Environment for Network Innovations (GENI) project, they are working to connect networked cloud sites as needed to provide resources and applications for experimentation and discovery, and production activities, including data-intensive computing.
The buzz phrase for their work is Networked Infrastructure as a Service (NIaaS), a model that allows for provisioning of network functions alongside compute and storage resource provisioning. The RENCI project is ExoGENI, a joint effort with Duke University and IBM that links NSF GENI resources to open cloud computing resources and dynamic circuit fabrics that allow users to specify the network services they need for their job in real time.
ExoGENI offers a glimpse of how research networks and the next version of the Internet could operate. You can read about it now in a new RENCI White Paper: Visions of a Future Internet: The ExoGENI Example.