AW: benchmark info

From: Norman Geist (norman.geist_at_uni-greifswald.de)
Date: Fri Feb 28 2014 - 01:18:46 CST

If you take the namd manual and lookup the commands in the apoa1 config file, you might figure out most of your questions. Usually the PDB holds the initial coordinates and some naming of atoms. In addition the PSF (if charmm FF is used) file will hold the atom/bind/angle/dihedral/.... parameters. Initial velocities are introduced randomly by the "temperature" keyword. Please notice that the "result" of a md simulation is _NOT_ _ONE_ set of new coordinates but a "trajectory" (DCD file) that can be analyzed afterwards. During simulation, restart information gets written to restartname.xsc/coor/vel if set, or othwerwise "outputname.restart.xsc/coor/vel":

*.xsc -> Box size
*.coor -> Coordinates
*.vel -> Velocities
*.dcd -> Trajectory

Additionally, one redirectes namds stdout to a logfile holding the energies and timing information. Please be aware that doing comparison benchmarks should come with some experience on both codes in case of getting the best performance configured out of them to be truly fair.

Additionally, as initial movements are introduced randomly you will usually not be able to reproduce a trajectory, this has nothing to do with precision. Also, if using parallelization, you will not be able reproduce the same energies/coordinates even with one code, starting with the same velocities. So you need to carefully describe your metric of "precision".

Norman Geist.

> -----Urspr√ľngliche Nachricht-----
> Von: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] Im
> Auftrag von Hassan Kianinejad
> Gesendet: Donnerstag, 27. Februar 2014 23:34
> An: namd-l_at_ks.uiuc.edu
> Betreff: namd-l: benchmark info
>
> Hi
>
> I am a graduate student researcher at UCLA Computer Science department
> and I have some questions about input and output files for testing your
> source code.
> I downloaded this benchmark: apoa1 and in this directory I looked at
> some of the files; however I did not quite figure out which file
> contains atom's info (like position; initial velocity; type; etc). And
> also what is included in the simulation output file? (should be final
> positions and velocity of all atoms, right?)
> Because I need to run my own source code to do the simulation
> sequentially and compare my own results to look for
> difference/precision.
>
> I would really appreciate it if you could guide me through this
> process.
>
> Kind regards,
>
> Hassan Kianinejad
> Graduate Student
> VLSI CAD Lab, UCLA

---
Diese E-Mail ist frei von Viren und Malware, denn der avast! Antivirus Schutz ist aktiv.
http://www.avast.com

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2015 - 23:20:33 CST