X, named by its developers for its design goal of 10 teraflops, ranks below Japan's Earth Simulator and ASCI Q, the Los Alamos National Laboratory's dedicated weapons computer. Jack Dongarra of the University of Tennessee, who also holds an appointment at Oak Ridge National Laboratory, compiled the listing, made available on a web site, http://www.
In terms of the speed of Virginia Tech's supercomputer, it can theoretically handle a potential 17 teraflops, or 17 trillion operations per second. The recorded 10.28 teraflops is "truly amazing" when one considers the machine was essentially put together in two months of construction time, says Jason Lockhart, one of the members of Virginia Tech's project team.
Dongarra, interviewed earlier by the Richmond Times Dispatch, said the "notable aspect" of Virginia Tech's supercomputer "is the $5.2 million price for all that computing power."
Experts place the cost of Japan's supercomputer at a range of $350 million to $400 million with an upwards figure of $1 billion when support for the entire resource is added. Its speed is about 35 teraflops.
As the cluster was being built, the university named Srinidhi Varadarajan, an assistant professor of computer science in the College of Engineering, the director of Tech's new Terascale Computing Facility. Lockhart, also of the College of Engineering, and Kevin Shinbaugh of information technology were named associate directors. Glenda Scales, assistant dean for research computing and distance learning, in the College of Engineering, and Pat Arvin, associate vice president for information technology, provided the overall direction for the project.
Varadarajan started the Virginia Tech initiative with a National Science Foundation grant to expand and upgrade a small supercomputer he was directing on the university campus. Conversations with Lockhart and others led to the grander goal.
"We believed that we could build a very high performance machine for a fifth to a tenth of the cost of what supercomputers now cost, and we did," said Hassan Aref, dean of Virginia Tech's College of Engineering and a former chief scientist at the San Diego Supercomputer. "And we wanted to have our own supercomputer to use for our new Institute for Critical Technology and Applied Science (ICTAS) where we will be conducting multidisciplinary work on such topics as nanoelectronics, aerodynamics and the molecular modeling of proteins. With this machine, our researchers will be able to build computer modeling in days, not years."
Virginia Tech teamed with Apple, Mellanox Technologies, Liebert, a division of Emerson Network Power, and Cisco to create X. They selected Apple's new Power Mac G5 running Mac OS X as the framework for the cluster. Mellanox supplied the primary communications fabric, drivers, cards and switches for the project. Cisco's Gigabit Ethernet switches were used for the secondary communications fabric to interconnect the cluster. Liebert's high-density rack mounted cooling system and computer racks along with power distribution equipment were specifically designed for the cluster.
"Apple is thrilled to be the framework for Virginia Tech's cluster," said Jon Rubinstein, Apple's senior vice president of hardware engineering. "The recognition of the cluster by The Top500 Project as the world's third fastest is a wonderful testament to the Virginia Tech team and the revolutionary Power Mac G5."