Georgia Tech’s PCMOS Chip Harnesses Noise and Utilizes It to Extract Energy Savings

Our editors review and recommend products to help you buy the stuff you need. If you make a purchase by clicking one of our links, we may earn a small share of the revenue. Our picks and opinions are independent from any business incentives.

Crest-LogoSM.jpg
March 14, 2006 – A research team from Georgia Tech has found a way to harness noise and use it to extract energy savings with its Probabilistic CMOS, or PCMOS, chip. The chip can save up to 500x more power than traditional CMOS technology. With digital cameras adding larger LCD screens and power-hungry lengthy lenses, better energy efficiency is needed. More and more cameras are adding high ISO sensitivities as well, so noise is quickly becoming a hot topic – and is unfortunately more readily available. While the PCMOS chip doesn’t suppress the noise, it at least uses it for a good cause.

"Probabilistic architectures extend PCMOS to computing substrates beyond devices," said Krishna Palem, head of the research team and founding director of the Center for Research in Embedded Systems & Technology, in a March 6 press release. "By mixing chip measurements and simulations, gains have been shown using this technology for such applications as Hyper-encryption as applied to computer security, and through cognitive applications such as speech recognition and pattern recognition as well as image decompression. The gains ranged from a factor of 10 to a factor of more than 500 over conventional architectural approaches."

The "conventional architectural approaches" include standard CMOS chips used in some digital cameras. When PCMOS and CMOS are compared side by side, some simulations showed the PCMOS was up to 560x more efficient. The research team used fake noise manufactured at the 0.25-micron level and used probability to save some battery power. This does not eliminate noise from the image, but according to the Georgia Tech press release, this is not a problem.

"Given the human ability to average [noise] routinely such as in voice when using cell phones, or in images when they are streamed to hand held devices, the user does not often notice the distortion as significant and is willing to pay the price for significant energy savings," the release read. Will users opt for clearer pictures or amazing battery life? That is still down the road, as it oftentimes takes years to incorporate fresh technology into consumer applications.

The Georgia Tech team presented their work at the Design, Automation and Test in Europe (DATE) Conference last week in Germany. Their paper can be found at http://www.crest.gatech.edu/palempbitscurrent/date2006.pdf.

Our editors review and recommend products to help you buy the stuff you need. If you make a purchase by clicking one of our links, we may earn a small share of the revenue. Our picks and opinions are independent from any business incentives.

What's Your Take?

All Comments
Our editors review and recommend products to help you buy the stuff you need. If you make a purchase by clicking one of our links, we may earn a small share of the revenue. Our picks and opinions are independent from any business incentives.
Advertisement - Continue Reading Below
Advertisement - Continue Reading Below
Advertisement - Continue Reading Below
Advertisement - Continue Reading Below
Advertisement - Continue Reading Below
Advertisement - Continue Reading Below