Brad Wheeler, CIO and vice president of IT at Indiana University, thinks that offering supercomputing as a shared utility to companies makes a lot of sense because it can provide them with standardized software packages as well as the means to host their application code and help in parallelizing it. The initial steps involve corporate legal counsels checking to ensure that proprietary research will be safe on a shared system. But once that bridge is crossed, "you start to see a lot more acceleration in its use," Wheeler said.
The role that HPC technology can play in developing new economic opportunities was illustrated in August, when Louisiana State University and the Louisiana government announced an agreement to open a quality assurance center with Electronic Arts Inc., a Redwood City, Calif.-based developer of computer games and other interactive entertainment software. The announcement followed the development of a digital media academic program that includes increased research in visualization on HPC systems.
Stacey Simmons, associate director of economic development at LSU's Center for Computation & Technology, said the state wants to build a visual-media economy. "We really wanted something that was an economic engine," Simmons said. But first, it was important to develop a computing, software and networking environment that can support that kind of economic activity, she added.
offers another way to gain access to HPC resources, and researchers at Rice University in Houston have used Amazon.com Inc.'s Elastic Compute Cloud, or EC2, service. But access to computing capacity is only one aspect of the problem. Applications and codes have to be adapted to run in parallel environments, and there is a need for people with the skills to make that happen, said Charles Koelbel, a research scientist at Rice.
The university is trying to make training in parallel programming as affordable and accessible as possible. As part of that effort, Rice is developing books that can be downloaded online, partly through a competition that challenges people to write about various parallel computing topics. A number of companies are backing the contest, including Chevron Corp., Sun Microsystems Inc. and Nvidia Corp. "These firms really need to have good people to help them do scientific computing," Koelbel said.