[gmx-users] P-LINCS and nstlist

[Mark Abraham]

On 19/05/2011 1:39 AM, Justin A. Lemkul wrote:
>
>
> Sławomir Stachura wrote:
>> Dear Users,
>> I am  not sure if I understood properly  how P-LINCS should 
>> parametrized for simulation. In manual it's written that "one 
>> coordinate communication step is required for the initial LINCS step 
>> and one for each iteration"; so as coordinate communication is 
>> tightly binded to neighbor searching, does it mean that nstlist 
>> should equal "1"? I mean - should neighbor list be updated every 
>> simulation step for proper use of constraints in parallel with LINCS?
>> I have searched for the answers in mailing list, but I was 
>> unsuccesful in finding tone.
>
> The quoted passage is related to domain decomposition communication, 
> not neighbor searching.  The algorithms involved are rather complex, 
> and (usually) work without much manual intervention.
>
> nstlist is more directly related to dt, and is often connected to the 
> force field, as well.  Values of 5-10 are standard.

Values of 10 with dt=0.002 are normal, but with longer time steps 
smaller values of nstlist are wise. constant = nstlist * dt is probably 
a wise rule of thumb.

Mark

>   Setting nstlist = 1 is usually only necessary for EM, not MD.  
> Excessive updating of the neighbor list can result in performance 
> loss, I believe.
>
> -Justin
>
>> Thanks in advance,
>>     Slawek-- 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
>>
>