[QE-developers] [QE-user] Speeding up QE calculations

[Natalia Kireeva]

Hello, 

You can try CPU overclocking. It gives you something for the processors with unlocked multiplier (processors with K index at the end - e.g. "Core i7-875K” ). Please, don’t chuckle, it seems it helps )) 
Be careful with kinetic energy cutoff options (in the papers you can find these parameters in the eV while in QE the Ry are used).
You can use the special tool in QE for the conversion of cif to QE input file/coordinates (in PW/tools) to obtain the reduced system. 

Best regards, 
NK

Natalia KIREEVA, PhD,
Frumkin Institute of Physical Chemistry and Electrochemistry RAS
tel. (+7) 916-8257704
e-mail: nkireeva at gmail.com; kireeva at phyche.ac.ru


> On 6 Jun 2020, at 17:18, Yuvam Bhateja <yuvamb16 at gmail.com> wrote:
> 
> Hey everyone,
> 
> I want to simulate my system consist of more than 200-300 atoms and will require performing VC relax, relax, etc, all types of calculation.
> I am using a cluster with 8 core CPU, 42 GB RAM (RAM can be increased but CPU cores cannot) and no GPU.
> I tried GPU acceleration but for some other reason it didn't work.
> Can someone suggest how to speedup my calculations? 
> Any tips and hits will be appreciated.
> 
> Regards
> Yuvam Bhateja
> _______________________________________________
> developers mailing list
> developers at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/developers




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20200606/6e428bb1/attachment.html>