Pittsburgh Research scientists are poised to take advantage of the enormous capacity of unused fiber that's lying underground waiting to be "lit." This so-called dark fiber, the result of the exuberant buildout of fiber optics in the 1990s, is available to anyone who needs to transfer massive amounts of data over long distances.
Interested researchers should contact Tom West. He's president and CEO of National LambdaRail Inc., a consortium of leading U.S. universities and private companies that is deploying a national networking infrastructure to advance networking research and, in the process, enable next-generation network applications in science, engineering and medicine.
"NLR aims to re-energize innovative research and development into next-generation network technologies, protocols, services and applications," West said.
West has been a force in U.S. national research and education networking for nearly two decades, including involvement with the National Science Foundation's NSFnet in the mid-1980s during the original round of Internet development. Before NLR, he headed the Corporation for Education Network Initiatives in California (Cenic), working to advance California's research and education network. West was CEO of both Cenic and NLR for 10 months until deciding to devote all his time to NLR.
Cenic is a not-for-profit corporation serving the California Institute of Technology, California State University, Stanford University, the University of California, the University of Southern California, the state's community colleges and its K-12 school system. John Silvester, Cenic chairman and NLR board member, described West's achievements: "It was the Cenic initiative led by Tom of deploying our next-generation CalRen network backbone, which utilizes leased dark fiber for a multilayer advanced network, that inspired the formation of NLR." The Cenic approach has been "emulated by many other states and even some other nations," Silvester said, and NLR "is constructing a similar infrastructure at the national level."
NLR has a five-year plan and $115 million, with $85 million coming from NLR members and corporate participants. It receives no government funding. "It's important that members fund first in order to attract top government agencies such as NASA to partake in it," West said. "Our goal is to have 40 networks around the country as part of NLR."
On Nov. 18, 2003, NLR announced that it had successfully taken the first step in its planned national network, lighting a path between StarLight, a switch/router facility in Chicago developed by universities and labs from around the world, and the Pittsburgh Supercomputing Center, which is backed by Carnegie-Mellon University, the University of Pittsburgh and Westinghouse Electric Co. By next spring, NLR expects to have the infrastructure in place for a network that spans the United States.
The infrastructure will involve individual "lambdas" for university researchers. A lambda is a lightwave that uses optical dense wavelength division multiplexing (DWDM), which can simultaneously transmit up to 40 wavelengths, each capable of transmitting 10 Gbits/second.
Cisco Systems Inc. is providing optical DWDM multiplexers, Ethernet switches and IP routers for the NLR network. Level 3 Communications Inc. and WilTel Communications provided the fiber.
West explained how NLR provides an infrastructure for many networks to exist side by side in the same fiber-optic cable pair. The networks will be physically and operationally independent, since each will be supported by its own lightwave.
In a keynote speech at SC2004 earlier this month, West called for network "plumbers" to support the NLR infrastructure at all levels to create a national-scale network.
"The U.S is late to this game," he said. "However, other countries that were more cautious in their fiber buildout while the U.S. forged ahead in the '90s are now finding they don't have the dark fiber available to them as we do. NLR strives to again stimulate and support innovative network research to go above and beyond the current incremental evolution of the Internet."
The results of such endeavors are expected to facilitate further commercial development and creation of new technologies and markets, thereby stimulating economic development and contributing to U.S. national competitiveness, according to West.
A holistic set of facilities is being provided to various research constituencies. And a few 10-Gbit lambdas are already owned and operated by research agencies, including Pacific Wave, UltraScienceNet and OptiPuter.
The principal investigator of the OptiPuter project is Larry Smarr, director of the California Institute for Telecommunications and Information Technology, or Cal-(IT)2, and an Internet applications trailblazer. The OptiPuter is an IP-based optical network that applies visualization technologies to tightly couple computational resources for visualizing large data sets remotely.
"There are applications where you cannot analyze individual small slices of an entire database, and you need to visualize the entire database," said Smarr. One such example is a database set of a rat's brain at the University of Illinois, Chicago, where the entire brain is visualized over 54 tile displays, all powered by AMD Opteron 64 processors.
Architecturally, the OptiPuter has optical networking as the central processing element, not individual computers. The resultant "supernetworks" enable scientists who are generating terabytes and petabytes of data to visualize, analyze and correlate their data interactively from multiple storage sites connected to optical nets.
Smarr is well-known in the network research community. For two decades, he conducted observational, theoretical and computational research in relativistic astrophysics. In 1985, he was named the founding director of the National Center for Supercomputing Applications (NCSA), based at the University of Illinois. In 2000, he became the founding director of Cal-(IT)2, which brings together faculty from the University of California campuses at San Diego and Irvine with industrial partners to explore future development of the Internet. Smarr was a member of the President's Information Technology Advisory Committee and is currently on the Advisory Committee to the NIH Director as well as the NASA Advisory Council.
He is now on a "collaboratory" kick. "The goal of a collaboratory is to eliminate distance between collaborating scientists and remote scientific instruments, distributed data repositories and other resources," he said. He is leading the multi-institutional, NSF-funded OptiPuter activity to create environments for high-performance science using lambdas across the campus, state, nation and globe whose entire bandwidth can be dedicated to a single campus researcher. In the United States, the backbone is the National LambdaRail, which is linked to the international Global Lambda Integrated Facility at the University of Amsterdam.
"The OptiPuter project aims at learning how to 'waste' bandwidth and storage in order to conserve 'scarce' computing in this new world of inverted values," said Smarr. Essentially, the OptiPuter is a "virtual" parallel computer in which the individual "processors" are widely distributed clusters. Research teams in southern California and Chicago are prototyping the OptiPuter initially on campus, metropolitan and state-wide optical fiber networks.
"We have to start thinking that access to a high-bandwidth network for research purposes should be as easy as turning on a light switch or picking up a phone," said NLR's West. "Our work as well as the OptiPuter effort will make that so."