Friday last week saw the official inauguration of a new research center in Aarhus that will conduct fundamental work on creating smart algorithms to enable massive amounts of computerised data to be accessed and processed with less memory power and energy than is currently the norm, on diverse computing platforms. The Center for Massive Data Algorithmics (Madalgo) has received DKK 30 m (USD 5.5 m) in support from the Danish National Research Foundation and a matching sum from the University of Aarhus.
Researchers at the new center in Aarhus will collaborate closely with two world leading players in the field - Massachusetts Institute of Technology (MIT) in the US and Germany's Max Planck Institut für Informatik (MPII). Professor Lars Arge, who heads up the Madalgo Center, comments: "Together we can solve difficult problems quickly, and find the way to make computers and machines work smarter."
Collaborating with Lars Arge and an associate professor in Aarhus, will be two associate professors at MIT and a professor and senior researcher at MPII, supported by a team of postgraduates and PhD students. The Madalgo Center will establish a vibrant international research environment in smart algorithms at the main site of the University of Aarhus.
Smart algorithms have undoubted potential, not least in the development of high capability features for small, portable devices with modest power resources, like digital cameras and mobile phones. A good example of this is manipulating digital images like photographs, which normally requires a computer with a fast processor and a substantial amount of memory. The reason why a digital camera or camera phone can handle the same task with relative ease is because it is programmed with smart algorithms.
Another area of powerful potential for smart algorithms is in opening new avenues of performance enhancement for microchips as it becomes increasingly difficult to cram ever more transistors into a given space. The well known Moore's Law, which states that the number of transistors on a chip doubles every 18 months, has held true since it was first stated in 1965. But as transistor packing density increases, heat production increases, threatening a physical upper limit to microchip capabilities. Professor Arge believes that super-smart algorithms can turbocharge microchip performance through step changes in processing efficiency, thereby significantly boosting their capabilities to handle massive amounts of data.
The news was reported by national daily newspaper Berlingske Tidende and on the University of Aarhus website.