[gmx-users] opls-aa, implicit solvent and GPUs

Ehud Schreiber schreib at compugen.co.il
Sun Sep 25 14:49:53 CEST 2011

Hi again,

Quoting from the current Gromacs "GPUs" webpage:

Force Fields: Supported FF are Amber, CHARMM. GROMOS and OPLS-AA are not

Also, it states

Forcefields that do not use combination rules for Lennard-Jones
interactions are not supported yet.

And I don't know whether opls-aa is such a forcefield.

So is or isn't it currently possible to accelerate opls-aa on GPUs?

Looking at the "Roadmap/GROMACS 4.6" webpage, on the up side:
Major features (almost) ready to be merged into the 4.6 branch:
    New GPU non-bonded kernels which work with almost all features of

While on the downside:

Features currently not supported by the new GPU and SSE kernels:
    Implicit solvent (but this will still by supported on the GPU
through OpenMM)

All this led me to my original question.

Thanks for your time (and for updating the Roadmap webpage...),


Message: 2
Date: Thu, 22 Sep 2011 10:55:56 -0400
From: "Justin A. Lemkul" <jalemkul at vt.edu>
Subject: Re: [gmx-users] opls-aa, implicit solvent and GPUs
To: Discussion list for GROMACS users <gmx-users at gromacs.org>
Message-ID: <4E7B4C7C.1010409 at vt.edu>
Content-Type: text/plain; charset=windows-1252; format=flowed

Ehud Schreiber wrote:
> Hi,
> I'm currently utilizing gromacs using the opls-aa forcefield in an 
> implicit-solvent setting.
> Can any of the developers roughly predict in what version, and when, 
> will GPU acceleration be applicable for such work? I am afraid the 
> gromacs 4.6 roadmap webpage left me confused.

You can do these simulations on GPU with the current version, or several
of the 
previous releases.


> By the way, the main roadmap webpage is totally outdated and needs 
> updating and cleaning...
> Thanks,
> Ehud.


Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080



More information about the gromacs.org_gmx-users mailing list