Professor Developing Energy-Efficient Computing Systems to Trim Power Usage

No matter how closely you or a third-party software package endeavor to monitor your computer’s power usage, they inevitably use more than is necessary, and certainly more than your wallet would like come bill-paying time. A University of Wisconsin-Madison engineering professor poses an interesting solution to this problem: instead of monitoring your system’s power, why not go straight to the source by offering low-power systems?

Nam Sung Kim, assistant professor of electrical and computer engineering, notes that the cost of powering U.S. data centers, such as those that process Google keywords every time users conduct a search, skyrocketed from $15 billion to $30 billion in 2008. The burning of many fossil fuels is required to generate all the electricity required to power the data centers, resulting in harmful environmental impacts. Kim is busily developing two algorithms to reduce power consumption. The first: program machines that process computations more efficiently, such as displaying images that utilize fewer pixels while still maintaining picture quality.


Image via Hudson Library

Kim’s second tactic is to reduce wasted energy during computations. “To achieve this, Kim is trying to identify which sections, called blocks, of computer circuits can be turned off during certain functions,” reads a news announcement from University of Wisconsin-Madison. Disabling blocks when their use is not required would reduce power. Kim’s endeavors have already netted him a CAREER Award from the National Science Foundation (NSF), which honors fledgling faculty members who successfully integrate advanced research and education into projects. Along with the award, Kim will receive a $437,000 grant to aid in his efforts.