Purdue Cluster challenge team starts fast

December 2, 2008

Purdue’s Cluster Challenge team started off fast at the SuperComputing ’08 Conference in Austin, Texas, Monday evening (Nov. 17).

As in 694 gigaflops fast.

That’s the score the Purdue team posted with the unusual SiCortex supercomputer it is using in the competition. The score was posted running the HPC LINPACK benchmarking software, the first step for all the teams in the Cluster Challenge.

Purdue placed first in the benchmarking segment overall and had the second highest LINPACK score, behind only National Tsing Hua University of Taiwan, which posted a score of 703 gigaflops. A combined team from Indiana University and Dresden University in Germany was third with a score of 526 gigaflops. None of the other four teams topped 500.

Purdue's team consists of Andy Howard, a senior in electrical and computer engineering technology from West Lafayette; Alex Younts, a sophomore in computer science from West Lafayette; David King, a senior in electrical and computer engineering technology from Lafayette; Paul Willmann, a senior in computer technology from Carmel; and Ryan Weinschenk, a senior in electrical and computer engineering technology from Noblesville.

The students are enrolled in a high performance computing class taught by Jeffrey Evans, an assistant professor in the Department of Electrical and Computer Engineering Technology. Preston Smith, senior UNIX system administrator for Purdue's Rosen Center for Advanced Computing, is serving as team leader.

The team, working in shifts, is now in a 44-hour marathon running simulated data in a battery of scientific applications relevant to fields ranging from genetics to designing jet airplanes. The competition ends Wednesday.

The teams are judged on their benchmarking score, the amount of data they can process with applications in the time allotted and their overall presentation, preparation and knowledge.

SiCortex builds supercomputers designed to deliver high performance using large numbers of slower, energy-efficient processors, both in power consumption and the cooling they require. Among other things, the company uses a unique, very fast “interconnect fabric”—the wiring that links its processors for working in concert—offsetting the raw speed disadvantage.

The Rosen Center, the research and discovery arm of Information Technology at Purdue (ITaP), the University’s central information technology organization, installed a top-of-the line SiCortex machine in June to test it from both performance and energy-saving perspectives, making Purdue the first university to do so. The SC1458 the Cluster Challenge team is using is the next step down from the Rosen Center’s SC5832. The numbers in both cases refer to the maximum number of processors the machines can contain.

Writer: Greg Kline, (765) 494-8167, gkline@purdue.edu

Originally posted: December 2, 2008