"Pruning" Microchips Cuts Energy Consumption by 50%, Doubles Computing Speeds
Photo by jpockele via Flickr Creative Commons
Technology behind microchips is constantly improved to make them smaller, faster and more energy efficient. But it turns out that one of the best ways to boost this speed and power conservation is to allow the chip to be a little lazy. Rather than requiring a chip to perform exact calculations, researcher have found that "pruning" parts of an integrated circuit that are rarely used and letting the chips make good-enough calculations, leads to an overall improvement. The technique is called "probabilistic pruning," and by allowing chips an 8% error rate in certain calculations, they can cut energy consumption by 50%, double the speed, and decrease the size of the chips.
"I believe this is the first time someone has taken an integrated circuit and said, 'Let's get rid of the part that we don't need,'" said principal investigator Krishna Palem, the Ken and Audrey Kennedy Professor of Computing at Rice University in Houston, who holds a joint appointment at Nanyang Technological University (NTU) in Singapore. "What we've shown is that we can boost performance and cut energy use simultaneously if we prune the unnecessary portions of the digital application-specific integrated circuits that are typically used in hearing aids, cameras and other multimedia devices."
It is a rather counter-intuitive idea -- after all, aren't computers supposed to eliminate mistakes whenever possible? Yet, the idea overall makes sense. Get rid of what you don't really truly need and experience a dramatic improvement in overall performance.
Pruning is part of this "inexact hardware" research that shows power demands on microprocessors can be cut as the wiggle room for error in predetermined ares is increased -- the researchers can decide what calculations can close enough instead of perfect. The smaller chips with lower energy consumption is finding use in areas such as hearing aids. The team expects the technology to help hearing aid batteries run four to five times longer without sacrificing quality.
GigaOm reports, "Hearing aids aren't the only application where researchers want to deploy probabilistic computing; Om has written about how Lyric Semiconductor, a firm coming out of MIT, hopes to use probabilistic computing to help process big data even faster. The company has raised $20 million to help improve Flash memory, networking functions and big data processing through its technology. The catch for Lyric and for any company with probabilistic computing goals. though, is that it changes the way software is written and thought about, so the need or perceived benefits will have to be huge to justify programming changes or applications that adopt the specialized computing will be confined to niche spaces.
The pruning technique is still undergoing testing, but looks to be a promising breakthrough for smaller, faster, longer-running devices in hearing and sight devices and many other areas as well.
Follow Jaymi on Twitter for more stories like this
More on Computer Chip Technology
Secret To Self-Repairing Computer Chip Is "Graceful Degradation"
World's First Light-Powered Chip Can Lead To Self-Powering Touchscreen Devices
Brazilian Trees Implanted with Microchips For Forest Management
Bees Equiped With Microchips Help Explain Hive Declines