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

The ongoing hardware evolution exhibits an escalation in the number, as wellas in the heterogeneity, of computing resources. The pressure to maintainreasonable levels of performance and portability forces application developersto leave the traditional programming paradigms and explore alternativesolutions. PaStiX is a parallel sparse direct solver, based on a dynamicscheduler for modern hierarchical manycore architectures. In this paper, westudy the benefits and limits of replacing the highly specialized internalscheduler of the PaStiX solver with two generic runtime systems: PaRSEC andStarPU. The tasks graph of the factorization step is made available to the tworuntimes, providing them the opportunity to process and optimize its traversalin order to maximize the algorithm efficiency for the targeted hardwareplatform. A comparative study of the performance of the PaStiX solver on top ofits native internal scheduler, PaRSEC, and StarPU frameworks, on differentexecution environments, is performed. The analysis highlights that thesegeneric task-based runtimes achieve comparable results to theapplication-optimized embedded scheduler on homogeneous platforms. Furthermore,they are able to significantly speed up the solver on heterogeneousenvironments by taking advantage of the accelerators while hiding thecomplexity of their efficient manipulation from the programmer.

Taking advantage of hybrid systems for sparse direct solvers via task-based runtimes