Home arrow What's News arrow Cameron & Feng at Supercomputing Conference
Cameron & Feng at Supercomputing Conference PDF Print E-mail


Virginia Tech will be well-represented at this year's Supercomputing Conference (SC06) , the premier international conference on high-performance computing, networking and storage. Virginia Tech researchers will present 3 of the 54 technical papers accepted to the conference. These papers were peer reviewed and selected among the 239 paper submissions.

 Dr. Wu Feng Dr. Wu-chun Feng

Dr. Wu-chun Feng, an associate professor in computer science and electrical & computer engineering, has two technical papers at the conference. The first paper, in collaboration with the University of Utah and North Carolina State University, describes a large-scale genomic sequence-search effort that generates critical information to biologists, including evolutionary, structural, and functional relationships between every sequence in the largest available database.

His second paper, in collaboration with Los Alamos National Laboratory, proposes a next-generation network protocol that is based on intelligent end-system-aware adaptation. This paper demonstrates how to improve end-to-end network performance by a minimum of 25% through intelligent scheduling in a processor's operating system.

"We're pleased for the opportunity to present our research across a broad spectrum of high-performance computing" says Feng, who joined the department in January 2006. "We hope that reporting our results in such a prestigious venue will help ensure that our solutions are disseminated for use in tomorrow's high-performance computing systems."

  Dr. Kirk W. Cameron

Dr. Kirk W. Cameron, leader of the SCAPE laboratory and an associate professor in Computer Science, also has a technical paper that will be showcased at this year's SC06 conference. His work, in collaboration with his postdoctoral fellow Dr. Xizhou Feng, describes a new parallel implementation of the Bayesian phylogenetic inference method for DNA sequence data. The PBPI software identifies equivalent tree estimates almost 1500 times faster than the most widely-used, best-available sequential program.

Virginia Tech's increased presence at this year's SC06 conference is no accident. "The CS department has made a number of strategic hire's in the past several years to build upon our strength in HPC systems," says department chair Dennis Kafura. "These papers are evidence of our commitment to leveraging System X to produce quality CS research."