English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Scientific applications are often irregular and characterized by largecomputationally-intensive parallel loops. Dynamic loop scheduling (DLS)techniques improve the performance of computationally-intensive scientificapplications via load balancing of their execution on high-performancecomputing (HPC) systems. Identifying the most suitable choices of datadistribution strategies, system sizes, and DLS techniques which improve theperformance of a given application, requires intensive assessment and a largenumber of exploratory native experiments (using real applications on realsystems), which may not always be feasible or practical due to associated timeand costs. In such cases, simulative experiments are more appropriate forstudying the performance of applications. This motivates the question of Howrealistic are the simulations of executions of scientific applications usingDLS on HPC platforms? In the present work, a methodology is devised to answerthis question. It involves the experimental verification and analysis of theperformance of DLS in scientific applications. The proposed methodology isemployed for a computer vision application executing using four DLS techniqueson two different HPC plat- forms, both via native and simulative experiments.The evaluation and analysis of the native and simulative results indicate thatthe accuracy of the simulative experiments is strongly influenced by theapproach used to extract the computational effort of the application (FLOP- ortime-based), the choice of application model representation into simulation(data or task parallel), and the available HPC subsystem models in thesimulator (multi-core CPUs, memory hierarchy, and network topology). Theminimum and the maximum percent errors achieved between the native and thesimulative experiments are 0.95% and 8.03%, respectively.

Experimental Verification and Analysis of Dynamic Loop Scheduling in Scientific Applications