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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

In the exascale computing era, optimizing MPI collective performance inhigh-performance computing (HPC) applications is critical. Current algorithmsface performance degradation due to system call overhead, page faults, ordata-copy latency, affecting HPC applications' efficiency and scalability. Toaddress these issues, we propose PiP-MColl, a Process-in-Process-basedMulti-object Inter-process MPI Collective design that maximizes small messageMPI collective performance at scale. PiP-MColl features efficient multiplesender and receiver collective algorithms and leverages Process-in-Processshared memory techniques to eliminate unnecessary system call, page faultoverhead, and extra data copy, improving intra- and inter-node message rate andthroughput. Our design also boosts performance for larger messages, resultingin comprehensive improvement for various message sizes. Experimental resultsshow that PiP-MColl outperforms popular MPI libraries, including OpenMPI,MVAPICH2, and Intel MPI, by up to 4.6X for MPI collectives like MPI_Scatter andMPI_Allgather.

Accelerating MPI Collectives with Process-in-Process-based Multi-object Techniques