Re: [AMBER] Poor performance of REMD simulation

From: Adrian Roitberg <roitberg.ufl.edu>
Date: Wed, 7 Jun 2017 16:58:02 -0400

What is the sped you would get if you ran just ONE simple md, not remd?

Are you using sander or pmemd?

Adrian


On 6/7/17 4:54 PM, MOHD HOMAIDUR RAHMAN wrote:
> Dear Amber User's
>
> I am simulating a small peptide ( Trpcage peptide in water) through T-REMD.
>
> The system details given below
> 1) The total no of atoms = 6302 (Peptide = 20 amino acid, water = 2000)
> 2) Temperature range 290 to 600
> (I used this web server <http://folding.bmc.uu.se/remd/>to predict
> temperature of each replica)
> 3) Total no of replica = 64 and no of CPU = 64
> 4) Input file
>
> ​ &cntrl
> imin = 0, ntx = 5, irest = 1,
> nstlim = 500, dt=0.002, cut=10.0,
> ntpr = 1000, ntwx = 1000, ntwr = 1000, ioutfm = 1,
> ntc = 2, ntf = 2, ntxo = 2, iwrap=1,
> ntb = 1, tautp = 2.0, ig= RANDOM_NUMBER,
> tempi = XXXXX, temp0 = XXXXX, ntt =3, gamma_ln = 1,
> ntp = 0, numexchg = 15000, nmropt=1,
> &end
> &wt TYPE='END'
> &end
> &ewald
> EW_TYPE=1
> &end
> DISANG=trpcage_wat_leap.CHIR.dat
> ​
>
> ​Now I have two question
> 1) Is there a way to improve performance of this REMD simulation. I am
> getting only 0.11 ns/day on 64 CPU.
> ​
> | Current Timing Info
> | -------------------
> | Total steps : 2500000 | Completed : 2077000 | Remaining : 423000
> |
> | Average timings for last 1000 steps:
> | Elapsed(s) = 1562.38 Per Step(ms) = 1562.38
> | ns/day = 0.11 seconds/ns = 781189.58
> |
> | Average timings for all steps:
> | Elapsed(s) = 3256969.81 Per Step(ms) = 1568.11
> | ns/day = 0.11 seconds/ns = 784056.29
> |
> | Estimated time remaining: 184.3 hours.
>
>
> ​2) Could I run multiple systems to generates large no of trajectory. My
> objective for this simulation is study stability and folding and unfolding
> of Trpcage protein in water (Similar to
> http://www.pnas.org/content/105/46/17754.full ). ​
>
> ​Thank you very much for valuable suggestion
> Regards
> Rahman​
>
> '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
> *M*d* H*omaidur* R**ahman *( Research Scholar )
> Lab No. 510, Computational Biophysics Lab.
> Department of Biotechnology
> Indian Institute of Technology-Madras
> Chennai-600 036, India
> *Mobile No = +91- 7845991785*
> '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber

-- 
Dr. Adrian E. Roitberg
University of Florida Research Foundation Professor
Department of Chemistry
University of Florida
roitberg.ufl.edu
352-392-6972
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Jun 07 2017 - 14:00:02 PDT
Custom Search