A team of researchers led by the University of Illinois at Chicago and the University of California at San Diego are creating an IP-based computing system for scientists who need to visualize and analyze massive amounts of data stored in multiple locations connected via optical networks.
Dubbed the OptIPuter, the federally funded research project is focused on earth, ocean and biosciences applications, where researchers have large repositories of 2-D and 3-D images that need analysis.
“We are removing bandwidth as an obstacle in data-intensive sciences,’’ explains Maxine Brown, project manager for the OptIPuter and associate director of the electronic visualization lab at the University of Illinois at Chicago. “With our system, the network is a backplane, not a network.’’
Brown gave a briefing about the OptIPuter at the TeraGrid ’06 conference held in Indianapolis this week. The conference was co-sponsored by Indiana University and Purdue University.
Brown said scientists who need to visualize massive amounts of images need guaranteed bandwidth, guaranteed latency and guaranteed scheduling for their processing jobs.
To meet this need, the OptIPuter is a cluster of computers that acts as a giant storage, compute or visualization peripheral. Each node of each cluster is attached via 1Gpbs or 10Gbps Ethernet directly to a backplane.
The OptIPuter acts like a parallel computer, where the individual processors are actually clusters of machines, the memory is large data repositories and the peripherals are visualization displays. The OptIPuter’s motherboard uses standard IP delivered over multiple dedicated lambdas. The OptIPuter uses advanced middleware and application toolkits to manage processing across the clusters.
Currently, research teams in Illinois and California are prototyping the OptIPuter on campus, metropolitan and statewide optical networks.
Begun in 2002, the OptIPuter research project is funded under a five-year, $13.9 million National Science Foundation grant.
While the OptIPuter is focused on scientific applications, researchers anticipate applications of the technology for emergency response, homeland defense and health services.