Re: Yet another NAMD speed concerns

From: Axel Kohlmeyer (akohlmey_at_cmm.chem.upenn.edu)
Date: Sat Apr 25 2009 - 09:10:14 CDT

On Sat, 2009-04-25 at 13:44 +0400, DimitryASuplatov wrote:
> Hello,
>
> I am running a simulation of my protein in water (cell = 15 angstroms
> from the protein). System contains 110349 atoms total.
> I have compiled namd 2.6 on xeon infiniband cluster uusing amd64
> installation procedure. I`ve also used icc and ifort with -xSSSE3

if you want to get the fastest execution, _particularly_
using SSE, why don't you get the 2.7 beta?? using the compiler
SSE flags is almost always a waste of effort unless the code
uses compiler intrinsics explicitly. _no_ compiler flag can
replace a good algorithm and a resourceful and smart programmer.

> flag.
>
> My force field parameters are
> =====================
> exclude scaled1-4
> 1-4scaling 1.0
> cutoff 15.

why 15 angstrom???

> switching on
> switchdist 10.
>
>
> My integrator parameters are
> =====================
> timestep 2.0 ;# 2fs/step
> rigidBonds all ;# needed for 2fs steps
> nonbondedFreq 1
> fullElectFrequency 4
> stepspercycle 20
>
>
> I use
> ====
> langevin on
> langevinPiston on
>
> cellBasisVector1 108 0. 0.
> cellBasisVector2 0. 102 0.
> cellBasisVector3 0. 0 105.
> cellOrigin 63.87940979 13.3578186035 32.2524185181
>
> PME yes
> PMEGridSizeX 125
> PMEGridSizeY 125
> PMEGridSizeZ 125
>
> MY PROBLEM IS THAT MY SYSTEM IS TOO SLOW
> ======================================================
> Utilization of 256 x Xeon 3Gz CPUs requires 41 hours for 10 ns
> !!!!!!!!!!!!!!!!!!!

how fast is it with less CPUs? have you made a proper scaling test???
are you certain that you actually are using the infiniband network?

what kind of machine do you have _exactly_.
what kind of infiniband do you have and have you made
tests on the through put on that? i've seen quite a few
networks that were not set up correctly and thus performance
was abysmal when used with too many MPI tasks.

> ======================================================
>
> I consider my system to be of a typical size. 100k atoms fro 10 ns is
> normal in our days. But it is very hard calculate with NAMD even on
> modern clusters.

it is very hard for _you_. i am certain there are a
lot of people that think very differently.

> 1/ Is this speed normal for namd or did i do something wrong?

most likely the second.

> 2/ Can I use 4fs for timestep with this parameters?

not unless you mess with the particle masses. if you
only want "approximate" results, why don't you use gromacs?

> 3/ what do you do to shortcut your calculations?

i'd rather do correct simulations.

cheers,
   axel.

> Thanks
> SDA

-- 
=======================================================================
Axel Kohlmeyer   akohlmey_at_cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:52:40 CST