[gmx-users] testing Gromacs 4.6 (git version) on GPUs

[Susan Chacko]

Thank you! Yes, I am able to run the 'dhfr-solv-PME.bench' benchmark with
mdrun -s topol.tpr -testverlet 

and it seems to be using the GPUs as well:
---
2 GPUs detected:
  #0: NVIDIA Tesla M2050, compute cap.: 2.0, ECC: yes, stat: compatible
  #1: NVIDIA Tesla M2050, compute cap.: 2.0, ECC: yes, stat: compatible

2 GPUs auto-selected to be used for this run: #0, #1
---

Best,
Susan.

On Oct 31, 2012, at 12:15 PM, Roland Schulz wrote:

> Hi,
> 
> you can use the tpr files from the tgz if you use the non-implicit "solv"
> (preferable PME) test.
> 
> Roland
> 
> On Wed, Oct 31, 2012 at 10:37 AM, Susan Chacko <susanc at helix.nih.gov> wrote:
> 
>> 
>> Hi all,
>> 
>> I wanted to test the built-in support in Gromacs 4.6 (non-OpenMM) for GPUs.
>> I downloaded the latest git version and it built successfully, appears
>> to link the right CUDA libraries etc.
>> 
>> I tried testing it with the Gromacs GPU benchmark suite
>> (http://www.gromacs.org/@api/deki/files/128/=gromacs-gpubench-dhfr.tar.gz)
>> 
>> The output md.log file reports:
>> ----
>> […]
>> NOTE: GPU(s) found, but the current simulation can not use GPUs
>>     To use a GPU, set the mdp option: cutoff-scheme = Verlet
>>     (for quick performance testing you can use the -testverlet option)
>> ----
>> 
>> So I tried 'mdrun -s topol.tpr -testverlet' and got:
>> -------------------------------------------------------
>> Program mdrun, VERSION 4.6-dev
>> Source code file: /usr/local/src/gromacs/gromacs/src/kernel/runner.c,
>> line: 700
>> 
>> Fatal error:
>> Can only convert old tpr files to the Verlet cut-off scheme with 3D pbc
>> For more information and tips for troubleshooting, please check the GROMACS
>> website at http://www.gromacs.org/Documentation/Errors
>> -------------------------------------------------------
>> 
>> Does anyone have a test set that I can use to check the performance of
>> Gromacs 4.6 on GPUs? I'm more of a sysadmin than a computational chemist,
>> so I'm sort of reliant on the provided benchmark suites.
>> 
>> Any suggestions welcomed and appreciated,
>> Susan.
>> 
>> 
>> --
>> gmx-users mailing list    gmx-users at gromacs.org
>> http://lists.gromacs.org/mailman/listinfo/gmx-users
>> * Please search the archive at
>> http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
>> * Please don't post (un)subscribe requests to the list. Use the
>> www interface or send it to gmx-users-request at gromacs.org.
>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>> 
>> 
>> 
>> 
>> 
> 
> 
> -- 
> ORNL/UT Center for Molecular Biophysics cmb.ornl.gov
> 865-241-1537, ORNL PO BOX 2008 MS6309
> --
> gmx-users mailing list    gmx-users at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-users
> * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> * Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

Susan Chacko
Helix/Biowulf Staff