try pmemd, and also look closely at the timing breakdowns at the end
of the output file. it will tell you which parts, if any, are not
scaling well. it also will show communication overhead. perhaps the
nonbond calculation is scling fine, but other parts are not (Amdahl's
law)
On Wed, Aug 27, 2008 at 5:04 PM, Ilyas Yildirim
<yildirim.pas.rochester.edu> wrote:
> Dear All,
>
> I did a benchmarking to see the performance of a dual quad core system.
> The system I tested have 3359 atoms in it (an RNA solvated with Na+ ions,
> and ~1000 water molecules). I used sander.MPI in the simulations. The
> machine is a dual quad core system with Intel Xeon CPUs (E5430,
> 2.66GHz) and 16 GB RAM. I did 4-5 test simulations for each case and here
> are the results:
>
> 1 CPU X 1
> 2 CPU X 1.88
> 4 CPU X 3.49
> 8 CPU X 5.78
>
> I have to say, I am very disappointed with the results. I had the same
> question a year ago, and Ross Walker gave a really nice answer. I am
> surprised that we cannot, still, get a full performance due to the bad cpu
> architecture.
>
> Is it possible that if I use pmemd, I can increase the efficiency? In any
> case, just wanted to share my experience with you.
>
> Best regards,
>
>> Did any of you have experience with running sander in dell 2 quad core
>> systems? We are going to buy a system for the lab to run sander in
>> parallel, and I was wondering how the performances of DELL 2 quad core
>> systems are. Can we get a x8 performance when we use all the CPU for one
>> simulation? Thanks in advance for your answers.
>
> --
> Ilyas Yildirim
> ---------------------------------------------------------------
> = Department of Physics - =
> = University of Rochester - =
> = 585-275-6766 (office) - 585-267-5644 (home) =
> = http://www.pas.rochester.edu/~yildirim/ =
> ---------------------------------------------------------------
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber.scripps.edu
> To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
> to majordomo.scripps.edu
>
--
===================================================================
Carlos L. Simmerling, Ph.D.
Associate Professor Phone: (631) 632-1336
Center for Structural Biology Fax: (631) 632-1555
CMM Bldg, Room G80
Stony Brook University E-mail: carlos.simmerling.gmail.com
Stony Brook, NY 11794-5115 Web: http://comp.chem.sunysb.edu
===================================================================
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
to majordomo.scripps.edu
Received on Sun Aug 31 2008 - 06:07:17 PDT
