High-performance computing, networking, storage and analysis across scientific disciplines

EVENT: Computational scientists from the U.S. Department of Energy’s Brookhaven National Laboratory will present a talk, demonstrations, and 18 posters at the international supercomputing conference known as SC11 to highlight their capabilities in meeting high-volume data demands across a range of scientific fields.

WHEN: November 12-18, 2011

WHERE: Washington State Convention Center in Seattle, WA.

DETAILS: The SC11 conference is expected to attract as many as 11,000 attendees from academia, industry and government. The meeting continues a long and successful tradition of engaging the international community in high performance computing, networking, storage and analysis. SC11 will place an emphasis on bringing together communities to facilitate information exchange, discussions and new collaborations for research and education related to innovating high performance computing applications and advancing scientific discovery and scholarship.

Brookhaven Lab’s contributions include:

Presentation on a method to accommodate multiple and concurrent network reservation requests required for predictable and reliable data transfers between multiple locations around the world. The solution developed by the Brookhaven team constructs schedules that accommodate a significantly larger number of requests compared to other, seemingly efficient, methods.

End-to-End Network QoS via Scheduling of Flexible Resource Reservation Requests, Shushant Sharma, Computational Science Center, Brookhaven National Laboratory, Thursday, November 17, 2:30-3 p.m.

Demonstration of data transfer applications on 40-gigabit-per-second to 100-gigabit-per-second networks.

Dantong Yu, Computational Science Center, Brookhaven National Laboratory, Booth #2443: Mon., 11/14, 8:30 p.m.; Tues., 11/15, 2 p.m.; Wed., 11/16, 2 p.m.; and Thurs., 11/17, 10 a.m.

18 Posters and movies covering research done at Brookhaven Lab and Stony Brook University, including: computing for the Large Hadron Collider (LHC); computational biology; computational materials science and nanoscience; climate science; high-energy and nuclear physics (lattice quantum chromodynamics, or QCD); environmental science; smart grid modeling; astrophysics; computational fluid dynamics; and accelerator physics.

“To be at the leading edge of scientific discovery requires the most advanced computing, networking, data analysis, and visualization that is available today,” said Michael McGuigan, interim director of Brookhaven’s Computational Science Center. “Supercomputing 2011 is the world’s best stage on which to deliver that message, and to demonstrate how our resources — our staff and computing assets — are meeting those needs.”