NSF opens new round of grants for TeraGrid time

Supercomputing to become more super with transition to next phase

The National Science Foundation is accepting applications for grants of time on its TeraGrid high performance computing infrastructure. The one-year grants, which will run through June 2011, will be the first to participate in the in the transition to the next phase of TeraGrid, called eXtreme Digital.

Researchers have until April 15 to apply for remote access to a series of supercomputing systems tied together by an advanced network.

The transition to eXtreme Digital is intended to provide significantly enhanced high performance computing resources and data services, with capabilities beyond those of existing platforms. “While a competition for the management of XD is ongoing, the program solicitation requires the winning team to provide continuity of service between TeraGrid and XD,” the NSF solicitation reads.

TeraGrid is a partnership between NSF’s Office of Cyberinfrastructure and 11 government, educational and research facilities that make computing time available on 20 supercomputing platforms. It is supported by grid software and high performance network connections, and provides data storage and management resources, as well as access to the computers themselves. Total TeraGrid resources now exceed 2 petaflops (a petaflops is 1,000 trillion floating point operations per second) combined processing power and 60 petabytes of online and archival storage.

Although a number of individual supercomputers have broken the petaflop barrier and the fastest, Jaguar at the Oak Ridge National Laboratory, operates at 2.33 petaflops, TeraGrid allocations still provide access to high-powered computing resources not readily available at many institutions.

For the next round of awards, researchers can request time on 12 systems, including TeraGrid’s two largest, Ranger at the Texas Advanced Computing Center (TACC) and Kraken at the National Institute for Computational Sciences (NICS), and the newest, the Longhorn remote visualization and data analysis system at TACC. Another remote visualization and data analysis system at NICS, Nautilus, will go into operation in October. The National Center for Supercomputing Applications (NCSA) soon will deploy a highly parallel shared memory supercomputer called Ember. With a peak performance of 16 teraflops, Ember will double the performance of its predecessor, the five-year-old Cobalt system.

Three platforms will be decommissioned on March 31 and will not be available for the next round of resource grants. They are Big Red e1350 at Indiana University, the NCSA Mercury IA-64 Cluster and the Pittsburgh Supercomputing Center BigBen XT3. To facilitate the transition to the next phase, TeraGrid will provide transition documentation, training, consulting and application support services for researchers through June.

TeraGrid allocates more than 1 billion processor hours to researchers each year, and applications received by April 15 will be considered for awards will be available from July 1, 2010, through June 30, 2011. In March, the TeraGrid Resource Allocations Committee awarded 281 million processor hours and 1.25 petabytes of data storage to 133 research teams.

Applicants for the March allocation should use TeraGrid’s online submission system.

About the Author

William Jackson is a Maryland-based freelance writer.

inside gcn

  • Catalyst supercomputer at Lawrence Livermore National Lab

    Energy Department launches new HPC initiative

Reader Comments

Please post your comments here. Comments are moderated, so they may not appear immediately after submitting. We will not post comments that we consider abusive or off-topic.

Please type the letters/numbers you see above

More from 1105 Public Sector Media Group