Photos of Larryblakeley

(Contact Info: larry at larryblakeley dot com)

Important Note: You will need to click this icon to download the free needed to view most of the images on this Web site - just a couple of clicks and you're "good to go."

Overview: New Ways to Sort Data

World data is doubling every three years and is now measured in exabytesùa billion billion bytes. This much data will require computing platforms that can deal with terabyte-level (1,024 gigabytes) workloads. Intel is investigating three fundamental classes of processing capabilities in order to make meaningful use of this enormous sea of information: Recognition, Mining and Synthesis (RMS).

These three classes provide a framework for investigating and developing systems and software that can address tomorrow's computing workloads. This article explains RMS classifications, gives many examples of how they could be applied, and discusses the implications of RMS for the platform architectures of the future.

The Coming Era of Tera

Today, the world's data outstrips our ability to comprehend it, much less take maximum advantage of it. According to the How Much Information project (http://www.larryblakeley.com/Articles/how_much_info.htm) at the University of California at Berkeley, print, film, magnetic and optical storage media produced about 5 exabytes of new information in 2002. That's the equivalent of 37,000 new libraries with book collections the size of the Library of Congress (17 million books). Approximately 92 percent of the new information was on magnetic media, mostly on hard drives.

The information explosion is not slowing down either. It's speeding up. According to the Berkeley project, new information stored on paper, film, magnetic and optical media doubled just between 1999 and 2002. As more of the world's population comes onto the Internet and uses technologies such as instant messaging, digital photography, digital video and weblogs, the rate of data increase will continue to accelerate.

Looking ahead, Intel predicts the Era of Tera is coming quickly. This will be an age when people need teraflops (a trillion floating point operations per second) of computing power, terabits (a trillion bits) per second of communications bandwidth, and terabytes (1,024 gigabytes) of data storage to handle the information all around them. You can already buy terabyte hard drives.

To handle all this information, people will need systems that can help them understand and interpret data. Search engines will not be enough. Tapping the Web alone, today's search engines often turn up thousands of documents in a single search, but many with minimal relevance. What's more, 50 million new or changed Web pages are added every day. And it's not just text; it's videos, photos, and various other kinds of media. Digitalization has given us infinite ways to create, store and display information. Now we need computers to be able to "see" data the way we do, to look beyond the 0's and 1's and identify what is useful to us and assemble it for our review.

- "Recognition, Mining and Synthesis Moves Computers to the Era of Tera," Pradeep Dubey, Senior Principal Engineer, Manager of Innovative Platform Architecture , Microprocessor Technology Lab http://www.intel.com/research/mrl/, Intel Corporation http://www.intel.com/index.htm, Technology@Intel Magazine http://www.intel.com/technology/magazine/index.htm, Volume 3, Issue 1, February 2005 http://www.intel.com/technology/magazine/computing/recognition-mining-synthesis-0205.htm

Directory: http://www.larryblakeley.com/Articles/monthly_articles/

File Name: pradeep_dubey200502

Post Date: February 8, 2005 at 9:15 AM CST; 1510 GMT