• Smaller data sets are problematic because — Overheads involved with launching a kernel — GPU underutilization: lower thread count has lower tolerance for long memory access latencies as there are no "other threads" ready to run. • Not all algorithms benefit from parallelization — Trade off between "parallelization effort" and "performance gain". — If it takes a lot of effort to parallelize an algorithm, it might mean that the algorithm is has no inherent parallelism, (e.g. the aircraft climbing metric).


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Massively parallel processing for dynamic airspace configuration


    Beteiligte:
    Gallet, Bart (Autor:in)


    Erscheinungsdatum :

    01.10.2011


    Format / Umfang :

    594773 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Massively parallel processing for dynamic airspace configuration

    Gallet, Bart / Brinton, Chris | IEEE | 2011


    Dynamic Airspace Configuration

    Bloem, Michael J. | NTRS | 2014


    Dynamic Terminal Airspace Configuration

    Hall, W.D. / Churchill, A. / Hwang, I. et al. | British Library Conference Proceedings | 2013



    Examining Airspace Structural Components and Configuration Practices for Dynamic Airspace Configuration

    Lee, P. / Mercer, J. / Gore, B. et al. | British Library Conference Proceedings | 2008