<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN">
<html><body style='font-family: Verdana,Geneva,sans-serif'>
<p>Just to add, I'd go through the trouble of manual compiling and tweaking fftw3 and OpenBLAS. <br />OpenBLAS can be serial or threaded, and its threaded performance is not always optimal (a known issue). <br />Fftw3 (from my experience) compiles and runs super fine on Ryzen, but the overall performance gain from it is 10% tops. <br />If you are looking for a more crucial bottleneck - it is probably not purely at the FFT part.<br /><br />On a Xeon supercomputer, PW runs faster (like, twice as fast or more) when compiled with Intel compiler, compared to GCC (although maybe it is just that one particular cluster). <br />For some reason, I use 8 or 12 threads, I could be I ran some tests for it. <br />On AMD machine, it might be different, and yet again it calls for some compilation tweaking. <br />I'd made a few versions and benchmarked them. <br /><br />Finally, you might be getting memory issues. Try making a CPU-heavy task and a memory-heavy task, both to run for at least few minutes, and see how much gain you get when going from 1 to 32 cores. <br /><br />Best regards, and good luck!<br />Andrii Shyichuk, University of Wrocław.<br /><br />W dniu 2020-11-16 20:36, Carlo Nervi napisał:</p>
<blockquote type="cite" style="padding-left:5px; border-left:#1010ff 2px solid; margin-left:5px"><!-- html ignored --><!-- head ignored --><!-- meta ignored -->
<div dir="ltr">While compiling with gcc/gfortran use -march=native option.
<div>With Intel i guess it is -xHost</div>
<div>This will enable the CPU native instruction.</div>
<div>With Epyc I guess a lot depends on the libraries. With gcc I used openblas and fftw3. On Ryzen they work great.</div>
<div>HTH,</div>
<div>Carlo</div>
</div>
<br />
<div class="gmail_quote">
<div class="gmail_attr" dir="ltr">Il giorno lun 16 nov 2020 alle ore 18:24 Husak Michal <<a href="mailto:Michal.Husak@vscht.cz">Michal.Husak@vscht.cz</a>> ha scritto:</div>
<blockquote class="gmail_quote" style="margin: 0px 0px 0px 0.8ex; border-left: 1px solid #cccccc; padding-left: 1ex;">No. I do no use hyperthreading.<br /> I had checked both OpenMPI and OpenMP - in outup file ...<br /> It shows what I set by the launch and enviromental variable correctly.<br /> If I use more than 4 cores in any way, I get no speed up ...<br /> ________________________________<br /> From: users <<a href="mailto:users-bounces@lists.quantum-espresso.org">users-bounces@lists.quantum-espresso.org</a>> on behalf of Pietro Delugas <<a href="mailto:pdelugas@sissa.it">pdelugas@sissa.it</a>><br /> Sent: Monday, November 16, 2020 3:42:59 PM<br /> To: Quantum ESPRESSO users Forum<br /> Subject: Re: [QE-users] Sub optimal performance on 32 core AMD machine<br /> <br /> Hi<br /> Are you using hyperthreading ?<br /> If yes, you should try to use only 32 cores. Check the total number of processors at the beginning of the run.<br /> The program makes an intense usage of the cpus so that hyperthreading is usually inefficient.<br /> Pietro<br /> <br /> Sent from Mail<<a href="https://go.microsoft.com/fwlink/?LinkId=550986">https://go.microsoft.com/fwlink/?LinkId=550986</a>> for Windows 10<br /> <br /> From: Michal Husak<mailto:<a href="mailto:Michal.Husak@vscht.cz">Michal.Husak@vscht.cz</a>><br /> Sent: Monday, November 16, 2020 3:19 PM<br /> To: Quantum ESPRESSO users Forum<mailto:<a href="mailto:users@lists.quantum-espresso.org">users@lists.quantum-espresso.org</a>><br /> Subject: [QE-users] Sub optimal performance on 32 core AMD machine<br /> <br /> I<br /> <br /> _______________________________________________<br /> Quantum ESPRESSO is supported by MaX (<a href="http://www.max-centre.eu">www.max-centre.eu</a>)<br /> users mailing list <a href="mailto:users@lists.quantum-espresso.org">users@lists.quantum-espresso.org</a><br /> <a href="https://lists.quantum-espresso.org/mailman/listinfo/users">https://lists.quantum-espresso.org/mailman/listinfo/users</a></blockquote>
</div>
<br clear="all" />
<div> </div>
-- <br />
<div class="gmail_signature" dir="ltr">
<div dir="ltr">
<div>
<div dir="ltr">
<div>
<div dir="ltr">
<div>
<div dir="ltr">
<div>
<div dir="ltr">
<div>
<div dir="ltr">
<div>
<div dir="ltr">
<div>
<div dir="ltr">
<div>
<div dir="ltr">
<div>
<div dir="ltr">
<pre>------------------------------------------------------------
Prof. Carlo Nervi <a href="mailto:carlo.nervi@unito.it">carlo.nervi@unito.it</a> <a>Tel:+39</a> 0116707507/8
Fax: +39 0116707855 - Dipartimento di Chimica, via
P. Giuria 7, 10125 Torino, Italy. <a href="http://lem.ch.unito.it/">http://lem.ch.unito.it/</a></pre>
<pre style="white-space: pre-wrap;"><strong>ICCC2020 has been postponed at 2022</strong></pre>
<pre style="white-space: pre-wrap;">ICCC 2022 28 August - 2 September 2022, Rimini, Italy: <a style="color: #1155cc;" href="http://www.iccc2020.com/">http://www.iccc2020.com</a>
International Conference on Coordination Chemistry (ICCC 2022)</pre>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
<br />
<pre>_______________________________________________
Quantum ESPRESSO is supported by MaX (<a href="http://www.max-centre.eu">www.max-centre.eu</a>)
users mailing list <a href="mailto:users@lists.quantum-espresso.org">users@lists.quantum-espresso.org</a>
<a href="https://lists.quantum-espresso.org/mailman/listinfo/users">https://lists.quantum-espresso.org/mailman/listinfo/users</a>
</pre>
</blockquote>
<p> </p>
<div> </div>
</body></html>