Hi, <br /><br />I will share with you my little experience with this warning.<br /><br />I believe it is just a matter of how elpa was compiled not being tuned to the machine you are running the simulation.The <span><span>ELPA_KERNEL</span></span><em><span><span> </span></span></em>default is AUTO and somehow it didnt manage to choose one of the options to the target machine.<br /><br />I tested with elpa or scalapack some single points and geometry optimizations and, even with that warning, results were the same.<br /><br />If you are still not confident about that you can try two changes under the GLOBAL section:<br /><br />1. Choose <span><span>PREFERRED_DIAG_LIBRARY</span></span><em> SCALAPACK -<br /><br />2. </em>Change <span><span>ELPA_KERNEL</span></span><span><span> to GENERIC or </span></span><span><span>GENERIC_SIMPLE.<br /><br />In some tests I did, SCALAPACK or GENERIC were (very roughly) 0.8% slower than GENERIC_SIMPLE. <br /><br /><br />Kind regards</span></span><br /><br /><div class="gmail_quote"><div dir="auto" class="gmail_attr">Em quinta-feira, 29 de fevereiro de 2024 às 13:18:01 UTC-3, Max Nishibuchi escreveu:<br/></div><blockquote class="gmail_quote" style="margin: 0 0 0 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">Hello all, <div><br></div><div>I have been trying to perform a geometric optimization on a polycarbonate structure, and I repeatedly see the following warning appear in the output file despite the relaxation continuing. </div><div><br></div><div> *** WARNING in fm/cp_fm_elpa.F:522 :: Setting real_kernel for ELPA failed ***<br></div><div><br></div><div>I am running the simulation using the most recent generic openmpi Docker container (available <a href="https://hub.docker.com/r/cp2k/cp2k/tags/" target="_blank" rel="nofollow" data-saferedirecturl="https://www.google.com/url?hl=pt&q=https://hub.docker.com/r/cp2k/cp2k/tags/&source=gmail&ust=1709514481600000&usg=AOvVaw0k0bHVY7o-ZnVWQNkMfxlr">here</a>, 2024.1_openmpi_generic_psmp). I have been able to converge geometric optimizations in the past while this warning was still present. </div><div><br></div><div>Do I need to worry about this warning? It's not clear to me what role real_kernel plays in the simulation. I've attached my input file for the sake of replication. </div></blockquote></div>
<p></p>
-- <br />
You received this message because you are subscribed to the Google Groups "cp2k" group.<br />
To unsubscribe from this group and stop receiving emails from it, send an email to <a href="mailto:cp2k+unsubscribe@googlegroups.com">cp2k+unsubscribe@googlegroups.com</a>.<br />
To view this discussion on the web visit <a href="https://groups.google.com/d/msgid/cp2k/a0be00c5-a8ef-4761-af04-91ae7fb7b948n%40googlegroups.com?utm_medium=email&utm_source=footer">https://groups.google.com/d/msgid/cp2k/a0be00c5-a8ef-4761-af04-91ae7fb7b948n%40googlegroups.com</a>.<br />