Tech Professor Lands Grant To Make Computers
Energy Efficient
A New Mexico Tech computer science professor has been awarded a three-year $299,000 grant from the National Science Foundation to develop techniques and software to reduce a high-performance computer’s energy usage.
Assistant professor of computer science department, Xiao Qin believes results from his research will save millions of dollars a year in electricity costs.
“It involves fundamentally novel energy-saving techniques which can be transferred to embedded disk systems where power constraints are more severe than conventional disk systems,” Qin said.
Qin said his buffer-disk (BUD) architecture will be designed for use in the large computer systems utilizing hundreds or even thousands of disk drives.
“This applies to scientific computing, and any computing that requires parallel disks consisting of multiple disks in high-performance computing systems,” Qin said. “This would also include systems designed for video surveillance, digital libraries, radio astronomy and large financial corporations.
“27 percent of the total power consumption is storage, and my preliminary estimates show a 38 percent reduction in that consumption,” he said.
That translates to an overall reduction of 10 percent in a computing system.
“For large systems where an enormous amount of electricity is used this could be a savings of $50,000 a year,” Qin said.
A more important benefit is a broad impact on the environment.
“You would see a reduction in emissions of air pollutants, and a substantial CO2 reduction,” he said.
Qin will team with Tao Xie of San Diego State University and Peter Lichtner of Los Alamos National Laboratories.
