The first component builds a dedicated, very high-speed network that will enable WSU researchers to transmit and receive large amounts of data quickly, across the campus or around the world. Those researchers may include physicists who collaborate on international projects in particle, nuclear and astrophysics; computer scientists working on high-speed networking; and biomedical researchers sequencing genes.
The dedicated network, or "Science DMZ", as investigators call it, will be exclusive to scientific research and separate from the standard university network for day-to-day traffic. The new network is scheduled for completion around mid-2013.
Patrick Gossman, Ph.D., deputy chief information officer for Computing and Information Technology and a co-investigator for the grant, likened the Science DMZ to a dedicated lane on the freeway with a 700 mph speed limit.
The upgrade will use specialized monitoring equipment on each end of a connection to identify bottlenecks and help keep data moving. Bottlenecks are a huge problem for researchers, Patrick Gossman said, especially when they are moving images from telescopes, particle accelerator data, brain maps or DNA sequences.
"They need to move huge files here, where they have the computing power to analyze them, or send them to collaborators at other institutions", he stated. "This upgrade will greatly speed up our researchers' ability to work, moving their research forward."
The second project component will upgrade the local network infrastructure in the physics building. This component will provide the necessary on-off ramps to the new Science DMZ as well as speed up their daily work.
"If you spend three days doing a calculation, you don't want to spend another three days waiting for it to download", David Cinabro stated. "With the new connection, that download could take just minutes."
In order to receive the grant, for which only 33 other institutions around the country were successful, WSU had to demonstrate a science-based need. The university's case was bolstered by the work of physicists with national and international research entities like the Fermi National Accelerator Laboratory, Argonne National Laboratory, the European Organization for Nuclear Research (CERN), the Large Hadron Collider and the Large Synoptic Survey Telescope (LSST).
"Without a science underpinning as to why we need a better computer network infrastructure, we wouldn't have gotten this done", David Cinabro stated.
In addition to the science, Patrick Gossman said, NSF officials also were interested in Wayne State's proposal because it includes exploration of software-defined (virtual) networking, which allows a physical network to be split into multiple separate networks, and because of the Science DMZ.
The new dedicated network includes 10-gigabit-per-second links between four buildings on campus: physics, Computing and Information Technology, computer science in the College of Engineering, and the Applied Genomics Technology Center at the School of Medicine, a tenfold improvement over current capacity. The Science DMZ can be extended to other critical research areas on campus for little additional cost, investigators said.
The Science DMZ also includes a 10 Gbps link to StarLight, a Chicago-based switch/router facility for access to high-performance national and global networks.
"It's very important that we have this fast link to StarLight and other Tier 1 data sites, because they have fast links to other places", David Cinabro stated. "We need access to that data to do this science, but we didn't have very good connections, and that was an impediment."
The future of science will involve collaborations using big data sets, he said. For example, the LSST, which he is working on, will photograph the entire available sky from northern Chile every few nights and thus produce terabytes of data daily. Additionally, some of his colleagues who collaborate internationally are trying to arrange it so researchers around the world can do things with Wayne State equipment.
"Without a fast, dedicated network, that's not even possible to contemplate", David Cinabro stated. "This is a great way to enable us to do more science - not just more of what we're already doing, but new kinds of things - and collaborate more directly in data-intensive activities."
Investigators are pleased the NSF has recognized the value of WSU research and the potential for accelerating it through this grant award. Once the network improvements are in place, future grant proposals can involve more data-intensive research and the possible creation of a WSU data farm, where large amounts of enhanced data drawn from larger hubs could reside for local use.