DARPA recently funded a 2 year study of the technical challenges of trying to accelerate computational capability by 1000X in roughly half the time it took to do the last 1000X. Such challenges, if solved, would enhance our ability to build systems that operate across the "extremes" of computing -- from world-class supercomputers running highly numeric applications to embedded processors (including spaceborne avionics) with orders of magnitude more capability for real-time image and sensor processing.
This talk will leverage the original Exascale study, to focus on two of the challenges relevant to space computing, namely memory architectures and energy utilization. This will be amplified by two subsequent application-specific studies on where energy is consumed in such systems: Linpack on the "clean sheet of paper" supercomputer architecture developed in the Exascale study, and an onboard Mars rover class rock finding algorithm on future tiled architectures similar to Maestro. The results are consistent but perhaps unexpected. We have crossed a threshold where the real energy and power problems of the future are in the memory and interconnect - not the processing logic.
Bio
Dr. Peter Kogge currently holds the Ted McCourtney Chair of Computer Engineering at the University of Notre Dame, with research interests in highly scalable computer architectures and nano-technologies. Prior to that he was an IBM Fellow with IBM's Federal System where among other projects he oversaw the development of arguably the first parallel computer to fly in space on the Shuttle, and the world's first multi-core chip in 1993 - on a DRAM process. He is the author of 2 books, including the first on the now ubiquitous technique of pipelining, and holds over 30 patents. Applications of his PhD research led to what is now called the Kogge-Stone adder, the fastest known adder constructed out of fixed fanout gates. He was also the chairman of the DARPA working group that developed the Exascale report.
Document date May 16, 2010.