Presentation Open Access

TREX : an innovative view of HPC usage applied to Quantum Monte Carlo simulations

Scemama, Anthony; de Oliveira Castro, Pablo; Valensi, Cedric; Jalby, William

    TREX : an innovative view of HPC usage applied to Quantum Monte Carlo simulations                                                                                                         


    The TREX[1] European Center of Excellence focuses on high accuracy quantum
    mechanical methods, essential in many different fields of application such as
    new material design or photochemistry. Among these methods, Quantum Monte Carlo
    (QMC) approaches are particularly well adapted to exascale architectures.
    Our ambition is to the help the community take advantage of exascale machines
    through the use of our HPC software.

    We will review in the presentation progress along the three following axes:

    * TREXIO[2]: A common I/O library and file format for easily exchanging data between
    applications, facilitating high-throughput computing workflows,

    * QMCkl[3]: A library of computational kernels specific to QMC applications
    written together by QMC and HPC experts, taking advantage of both CPUs and GPUs,

    * An integrated workflow including performance analysis (MAQAO[4]) and numerical
    accuracy measurements (Verificarlo[5]) to be used for the development of QMC
    kernels and more generally for improving the applications. In particular, we
    plan to identify the best performance usage of QMCkl and also to adjust the
    performance with numerical precision requirements.

    ------------------------------------

    [1] https://trex-coe.eu
    [2] https://github.com/trex-coe/trexio
    [3] https://trex-coe.github.io/qmckl
    [4] https://www.maqao.org
    [5] https://github.com/verificarlo

    ------------------------------------
 

Files (6.8 MB)
Name Size
trex_isc.pdf
md5:df7a0950dabd5393b3c90795f5d1fd54
6.8 MB Download
16
39
views
downloads
All versions This version
Views 1616
Downloads 3939
Data volume 263.4 MB263.4 MB
Unique views 1414
Unique downloads 3535

Share

Cite as