Tutorial-l Mailing List
From: Dudas, Balint (balint.dudas_at_kcl.ac.uk)
Date: Thu Jun 14 2018 - 04:31:18 CDT
- Next message: Mina Ebrahimy: "ionize water box by add ions plugin"
- Previous message: Fotis Baltoumas: "Re: NAMD-QM/MM: "QM forces are not compatible with CUDA at this time""
- In reply to: Fotis Baltoumas: "Re: NAMD-QM/MM: "QM forces are not compatible with CUDA at this time""
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Hello,
Thanks for the answer it seems that it was the problem. Now I am just facing a new kind of problem, running this tutorial - in which there are only +-23 000 atoms I get an error message: "Memory allocation failed on processor 0". This was not the case when using the same setup for MM only...
Do you know what the problem might be?
________________________________
From: fbaltoumas_at_biol.uoa.gr <fbaltoumas_at_biol.uoa.gr> on behalf of Fotis Baltoumas <fbaltoumas_at_biol.uoa.gr>
Sent: 13 June 2018 11:09:35
To: Dudas, Balint; tutorial-l_at_ks.uiuc.edu
Subject: Re: tutorial-l: NAMD-QM/MM: "QM forces are not compatible with CUDA at this time"
Hello,
I think the error is pretty much self-explanatory. The QM/MM interface is not compatible with CUDA yet, therefore the CUDA-enabled version of NAMD will not run.
The matter can be easily resolved; simply compile a second NAMD version without CUDA support (i.e. using only CPUs) and use that. Or download one of the precompiled distributions if you're in a hurry.
Just a note, the version you say you have is not 2.12, it's a nightly build with newer features, that will eventually become version 2.13. So I would advise you to consult the nightly build's user guide for more information (http://www.ks.uiuc.edu/Research/namd/cvs/ug/ Fotis
On 13-Jun-18 12:42, Dudas, Balint wrote:
Dear All,
I'm trying to go through the 'NAMD-QM/MM' tutorial and have already got to the point where I prepared (with the help of QwikMD) the inputs for the QMMM-Min simulation. Once I load it into QwikMD and launch the 'Start Minimization' simulation (and select the number of CPUs to use and set up the namd2 executable path - which worked perfectly without the QMMM before) it returns with a FATAL ERROR message. In the .log file everything works until the point where QMMM is needed. There I get the message:
"FATAL ERROR: QM forces are not compatible with CUDA at this time".
I'm using a version of NAMD: NAMD_Git-2018-06-13_Linux-x86_64-multicore-CUDA (2.12). MOPAC has been installed and CUDA-5.5 has been set up for. The same error message occurs when using ORCA instead of MOPAC so I guess it is independent of the QM software applied.
Could anyone help what the problem might be?
--
Fotis A. Baltoumas, Bsc, Msc
Phd Candidate, Bioinformatics Postgraduate Programme
Section of Cell Biology and Biophysics
Department of Biology
National & Kapodistrian University of Athens
Panepistimiopolis, Athens 157 01, GREECE
--------------------------------------
email : fbaltoumas_at_biol.uoa.gr<mailto:fbaltoumas_at_biol.uoa.gr>
http://biophysics.biol.uoa.gr