<div dir="ltr"><div dir="ltr">On Fri, Feb 4, 2022 at 8:10 AM Pietro Bonfa' <<a href="mailto:pietro.bonfa@unipr.it">pietro.bonfa@unipr.it</a>> wrote:<br></div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><br>
Do not use openMP, use MPI.<br clear="all"></blockquote><div><br></div><div>well, one may use OpenMP as well, but <br></div><div>- it is typically less efficient than MPI parallelization,</div><div>- 32 threads per MPI process are too many, and<br></div><div>- support for OpenMP in old versions (6.4.1) is not great</div><div><br></div><div>Paolo<br></div></div>-- <br><div dir="ltr" class="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div>Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,<br>Univ. Udine, via delle Scienze 206, 33100 Udine, Italy<br>Phone +39-0432-558216, fax +39-0432-558222<br><br></div></div></div></div></div></div>