INTEL'S ITANIUM AND McKinley processors will be used in a distributed scientific computing system, dubbed TeraGrid, allowing researchers to analyze, simulate, and help solve complex scientific problems, Intel announced Thursday.

   ADVERTISEMENT
  

Free IT resource

Hear how top CIOs turn change into a competitive advantage.

Sponsored by HP

Free IT resource

Attend the SOA Executive Forum: Breaking SOA Bottlenecks SOAExecForum.com/may2007

Sponsored by InfoWorld

RELATED LINKS

IDG ENTERPRISE NETWORK
• More App Development News...  (ComputerWorld)
• JBoss buys former HP middleware  (ComputerWorld)

TOP NEWS 


IT SOLUTION SEARCH
The TeraGrid is a grid computing implementation that will serve as the underpinnings for what's been called the world's first multisite supercomputing system, the DTF (Distributed Terascale Facility). Grid computing is a relatively new way to exploit the Internet to connect processors in massively virtual systems.

The effort will be funded with $53 million from the National Science Foundation (NSF).

The DTF will perform 11.6 trillion calculations per second and store more than 450 trillion bytes of data via the comprehensive infrastructure of the TeraGrid, which will link computers, visualization systems, and data at four sites through a 40-billion bits-per-second optical network, according to the NSF.

The TeraGrid will help the NSF address complex scientific research, including molecular modeling for disease detection, cures, and drug discovery; automobile crash simulations; research on alternative energy sources; and climate and atmospheric simulations for more accurate weather predictions, Intel said.

The TeraGrid system will link more than 3,300 Itanium processors as well as have the ability to store, access, and share more than 450TB of information, the company said in a statement.

The TeraGrid system, which is expected to be completed next year, will consist of clustered servers from IBM running the Linux operating system and will be connected by a high-speed optical network from Qwest Communications International.

In addition to supplying more than 1,000 IBM eServers running Linux, IBM Global Services will do the integration work for the system, said Dan Powers, IBM's director of early stage Internet technology, in a conference call with press and analysts. The cluster will also use GPFS (General Parallel File System) software, Powers said.

The completed system will be the world's largest computer running Linux, Dan Reed, director of the National Center for Supercomputing Applications (NCSA), said in the conference call. TeraGrid will be capable of 8T flops [floating point operations per second], Reed said. "We expect that in time this will grow to 10T flops or beyond," he added.

The largest portion of the system's computing power will be at the NCSA at the University of Illinois in Urbana-Champaign. The NCSA has three partners that will participate in the project: the San Diego Supercomputer Center (SDSC) at the University of California, San Diego; the Argonne National Laboratory in suburban Chicago; and the California Institute of Technology in Pasadena.

"This will really eliminate the tyranny of time and distance," Reed said. "We believe it's the future of computing, and it will change the way we do science and engineering research."

Launched in servers and workstations last month, Itanium is Intel's first 64-bit microprocessor and is intended to help the company compete in high-end server markets where systems from the likes of Sun Microsystems and Hewlett-Packard dominate. McKinley will be the second chip in the Itanium family and is due early next year, according to Intel.