This paper describes a fully programmable 6 processor die that was implemented in IBM's 130 nanometer 8SF process. It is capable of operating as two triple voted processors each with 6 Mbytes of Embedded Dynamic Random Access Memory (EDRAM) or 6 independent processors each with 2 Mbytes of EDRAM. In triple vote mode the processor counts and recovers from single event upsets. It supports external Double Data Rate (DDR) Synchronous Dynamic Random Access Memory (SDRAM), and has two Spacewire ports, a 4 GBit input port, and a 4 Gbit output port. The processor core with memory performs approximately 2.5 Giga Floating Point Operations Per Second (GFLOPS) per Watt, with worst case input/output power its performance is approximately 1 GFLOPS per Watt. The processor's efficient messaging allows hundreds of processors to be applied to applications such as radar.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    High Performance Space Computing


    Contributors:


    Publication date :

    2007-03-01


    Size :

    725771 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    High performance space VPX payload computing architecture study

    Alena, Richard / Collier, Patrick / Ahkter, Mohammad et al. | IEEE | 2016


    High performance enabled space object tracking via cloud computing

    Wei, Sixiao / Jia, Bin / Chen, Zhijiang et al. | IEEE | 2018


    High-Performance Computing Challenges in Aeronautics and Space Applications

    Basermann, Achim | German Aerospace Center (DLR) | 2017

    Free access

    A reconfigurable computing board for high performance processing in space

    Van Buren, D. / Murray, P. / Langley, T. | IEEE | 2004