Re: Error in running simulation

From: René Hafner TUK (hamburge_at_physik.uni-kl.de)
Date: Mon Jul 26 2021 - 10:31:31 CDT

Hi

    your runtime heavily depends on the number of atoms in your system.

    What I strongly recommend is reducing your system size as much as
    possible (might require testing of different system but it depends
    if its worth the time) usage of a GPU if available or at least a
    cluster node with more than 8 cores (if available),

    Equilibration time does not depend on production runtime.

    Equilibration means you run such that the
    temperature/volume/pressure/energy/density have settled down and
    e.g. no major structural changes occur in your protein  (RMSD) and
    other quantities. All depending on the ensemble and your quantities
    of interest.

    Though depending on your system maybe 10-100ps are enough of
    equilibration, but maybe you even need a few ns.

Kind regards

René

On 7/26/2021 5:10 PM, Ruturaj warake wrote:
> Dear Users,
> Thanks for replying.
>
> One more thing,
> I used the command +p8. Still it's taking 180 seconds for
> equilibration and 195 seconds for production to run 5000 numsteps with
> 2 fs time step.
>
> While running NAMD,
> We need to run equilibration step and then production.
> Suppose if I want to run simulation for 100 ns then do I need to run
> equilibration for 5,00,00,000 steps and then again production for same
> steps ?
>
> What if I reduce the equilibration steps ?
> Please do help me out.
>
> Thanks and Regards
>
> Ruturaj Warake,
> Niper, Hyderabad,
> India
>
>
> On Mon, 26 Jul 2021, 19:00 Ruturaj warake, <ruturajwarake47_at_gmail.com
> <mailto:ruturajwarake47_at_gmail.com>> wrote:
>
> Dear users,
> I am running simulation on my laptop.
> I have activated all 8 cores of processor but when I run the
> simulation it shows 1 core processor is in use. How to bring all 8
> processors in use ?
> Please help for the same.
>
> Kindly reply to this mail.
>
> Regards,
> Mr. Ruturaj Warake
>

-- 
--
Dipl.-Phys. René Hafner
TU Kaiserslautern
Germany

This archive was generated by hypermail 2.1.6 : Fri Dec 31 2021 - 23:17:11 CST