Hi Azat,
> My question is about scaling of AMBER on multiCPU systems.
> I had info, that Amber can not utilise more number of CPUs than are
> inside
> one node, is it right?
> For example, was found that AMBER could use not more than 8 CPUs
> effectively
> when there were 8 CPUs per cluster node.
> Is this law working for any computer systems - as workstations as
> clusters?
If I understand you correctly you are asking for some kind of perpetual
motion machine / something for nothing right? Is this really what you mean?
I.e. you have 8 real cores in your machine but you would like to get speedup
as if you had 16 physical cores in that machine? - This is clearly not going
to work for many reasons, some related to fundamental laws of physics
regarding conservation of energy.
You might want to clarify what you mean by this question.
> Also was found that AMBER works awfully on Teragride clusters, on the
> same reasons.
> May you specify the best parameters and possible speed of MD run on
> next systems:
> multicore workstations; Teragrid cluster
Which teragrid clusters and what exactly are you trying to run. Performance
in parallel is very tightly coupled with the size of system being run and
the type of calculation. Typically bigger systems scale better.
This page needs updating for the newer teragrid machines but see this:
https://www.teragrid.org/web/user-support/amber_benchmark
> Additional question. There is possibility to use Amber parameter files
> with
> another MD simulation software, for example NAMD, etc. May you comment,
> how
> perfectly they utilise Amber parameters, and how quality results may be
> created by this way?
Yes you can do this. People have run the amber force field in other
programs. How accurately it is supported is a matter for debate but
certainly in cases where it has been done carefully, such as NAMD, one can
get machine level agreement between AMBER and NAMD. Although most of the
disagreements are user error, not setting the 1-4 scaling correctly etc.
Note for most systems AMBER (if you use pmemd) and NAMD performance is
roughly equivalent. E.g. for DHFR.
See the section entitled "Using the Amber force field in other software
packages" on the amber webpage.
The second item in the list for example:
http://www.ks.uiuc.edu/Research/namd/2.8b3/ug/node14.html
Although this may need updating. I know a couple of people were working on
getting machine level correspondence between AMBER and NAMD. I'll let others
comment here. If you do not hear any more you may want to post this as a
standalone question with a more relevant subject line. E.g. AMBER force
field in NAMD.
All the best
Ross
/\
\/
|\oss Walker
---------------------------------------------------------
| Assistant Research Professor |
| San Diego Supercomputer Center |
| Adjunct Assistant Professor |
| Dept. of Chemistry and Biochemistry |
| University of California San Diego |
| NVIDIA Fellow |
|
http://www.rosswalker.co.uk |
http://www.wmd-lab.org/ |
| Tel: +1 858 822 0854 | EMail:- ross.rosswalker.co.uk |
---------------------------------------------------------
Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue May 24 2011 - 19:00:02 PDT
