[QE-users] Hybrid functionals using GPU accelerated QE

[Pietro BONFA']

Dear Giovani,

what you write is correct, the last release of the GPU version of pw.x
does not include acceleration for the Hartree-Fock part.

There is however a working version available for you to try in a
separate development branch at this address:

https://gitlab.com/QEF/q-e-gpu/tree/gpu-exx

This will eventually be merged and made available in the next release of
the GPU code.

Kind regards,
Pietro Bonfa'


On 6/22/19 1:12 AM, Giovani Rech wrote:
> Dear QE community,
>
> I’ve been testing the GPU accelerated version of QE using a hybrid functional (pbe0 and vdw-df-c090, to be exact, pun intended) and noticed that the use of the GPUs are virtually zero while the Hartree-Fock part is being calculated. This brings me to conclude that the the exact exchange calculation is not yet implemented for GPU. Can anyone confirm this info? I’ve not been able to find this information in the readmes or release notes.
>
> Thank you,
>
> Giovani Rech
> Universidade de Caxias do Sul
> Caxias do Sul - RS, Brasil
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
>


--
Pietro Bonfà
Department of Mathematical, Physical and Computer Sciences,
University of Parma,
Italy

Firma il tuo 5 per mille all’Università di Parma e aiuta così i nostri studenti che vogliono realizzare un’esperienza di studio all’estero - Indica 00308780345 nella tua denuncia dei redditi.