Via HPCwire, a podcast piece from IEEE Spectrum from a couple weeks ago that I’ve just added to my iPhone for drive home listening. In it Peter Kogge talks about the DARPA report that throws water on the idea that we’ll get to exascale computing anytime soon (if ever):
The U.S. Defense Advanced Research Projects Agency (DARPA) asked a group of eminent engineers and computer scientists to figure out how to get a thousandfold increase in supercomputing ability by 2015. The group, led by University of Notre Dame professor and IEEE Fellow Peter M. Kogge, found that if technology trends continue into the future, exascale computing will still be just a dream in 2015. According to the report, getting to exascale by then—or ever—requires among other things an intensive focus on reducing the power devoted to moving data around in computers. However, some critics point out that a similar report in the mid-1990s—predicting the technologies needed for petascale computing—has now proved to be off the mark. And some say the DARPA group did not properly consider new types of applications supercomputers will be running in the future. Kogge explains the report’s findings on Spectrum Radio at http://www.spectrum.ieee.org/dec08/scompod.