The Supercomputing (SC) conference hosts the annual Student Cluster Competition to challenge university students to build the most powerful computing clusters within a 26-amp power limit. At SC11, which took place in Seattle, Wash. this week, four of the top five teams used NVIDIA GPUs to attain Student Cluster Competition glory.
The top teams received major recognition in the supercomputing industry, and all participating teams gained valuable experience in building computing clusters. The students’ clusters were benchmarked and scored using scientific applications selected by contest organizers.
 > |
| The National Tsing Hua University team’s cluster. |
The top five SC11 Student Cluster Competition teams were:
- Taiwan – National Tsing Hua University
- China – National University of Defense Technology
- USA – University of Texas
- USA – Boston University
- Russia – Nizhni Novgorod State University
By the way, don’t feel bad for the Russian team. Nizhni Novgorod State University took home the prize for the highest overall performance, as measured on LINPACK. The Russian team managed to squeeze almost 2 teraflops of performance out of their small, low-power cluster.
These accomplishments are remarkable, and we congratulate Team Taiwan on their victory. Kudos to all the student teams in this competition. Student competitions like this can only help accelerate the course of computational science.
If you’d like to learn more about the competition, I encourage you to check out the SC11 Student Cluster Competition web site.
Tagged: cluster, CUDA, GPGPU, gpu, SC11, supercomuting, Tesla
