Fwd: NAMD + PLUMED COMPILATION

From: Francesco Pietra (chiendarret_at_gmail.com)
Date: Fri Mar 01 2013 - 00:52:14 CST

Sorry, I forgot the list.

---------- Forwarded message ----------
From: Francesco Pietra <chiendarret_at_gmail.com>
Date: Fri, Mar 1, 2013 at 7:51 AM
Subject: Re: namd-l: NAMD + PLUMED COMPILATION
To: Giacomo Fiorin <giacomo.fiorin_at_gmail.com>

Giacomo:
Sorry if I appeared to be snippy. It was not what I wanted.

I am first trying on a simple machine, where namd 2.9 cuda runs without problem:

Gigabyte X79-UD3
Two N680GTX
Intel i7-39330K
Two HDs raid 1 1Tera each
Debian amd64 wheezy

What I was asking was about the config stage of namd compilations (the
question does not depend from the presence or not of plumed), as, so
far, I used the binary provided by namd:

 ./config Linux-x86_64-g++ --charm-arch multicore-linux64 --with-cuda

i.e., whether the multicore flag should also be used. That was not
clear to me from instructions provided by namd. However, at this
point, I agree with you that, instead of posting the question, I could
have tried the various options. As to plumed, I had already compiled
it, after charm++.

Thanks
francesco pietra

On Thu, Feb 28, 2013 at 6:26 PM, Giacomo Fiorin
<giacomo.fiorin_at_gmail.com> wrote:
> Francesco: please don't be snippy. You did not give any information about
> your machine, like most computational scientists would tell you. I think
> you would find useful posts on this subject in the VMD and LAMMPS mailing
> lists.
>
> The PLUMED patch (as its documentation says) is applied to the source code
> before compilation, and doesn't affect the NAMD Makefiles.
>
> Just apply the patch, and compile the patched NAMD as you would the vanilla
> NAMD.
>
> Still in doubt but unwilling to provide details? Try the first compilation
> option that will succeed, and run some test calculations on the damn thing.
>
>
>
> On Thu, Feb 28, 2013 at 12:02 PM, Francesco Pietra <chiendarret_at_gmail.com>
> wrote:
>>
>> Sorry, previously I forgot the subject
>>
>>
>> ---------- Forwarded message ----------
>> From: Francesco Pietra <chiendarret_at_gmail.com>
>> Date: Thu, Feb 28, 2013 at 5:59 PM
>> Subject:
>> To: NAMD <namd-l_at_ks.uiuc.edu>, giacomo.fiorin_at_gmail.com
>>
>>
>> Hi:
>> Probably this message of mine, about the compilation of NAMD with
>> PLUMED included, has gone submerged, and thus unnoticed (or it is such
>> a naive question that does not warrant a reply):
>>
>> Thanks of the advice. A doubt about the last stage, when I have
>> everything (namd code, tcl fftw, plumed with its env var) under the
>> namd directory,
>>
>> export plumedir="root_of_plumed_package"
>> cp $plumedir/patches/plumedpatch_namd_x.y.sh
>> ./config Linux-x86_64-g++ --charm-arch multicore-linux64 --with-cuda
>> ./plumedpatch_namd_2.9.sh -patch
>> make
>>
>> is "--charm-arch multicore-linux64 --with-cuda" correct, or just
>> "--with-cuda" only?
>>
>> thanks francesco
>>
>

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:20:59 CST