FEP failure - FFTW problem?

From: JC Gumbart (gumbart_at_ks.uiuc.edu)
Date: Fri Sep 20 2013 - 14:16:37 CDT

• Next message: Vidhya Sankar: "About Muticore Linux Ibnstalltion if NAmd"
• Previous message: Francesco Pietra: "Re: About the scope of show_replicas.vmd"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

I've run into a strange issue trying to run FEP on a new (computer) system. So here is what I expect to get at the beginning of the simulation in my *fepout file:

> # STEP Elec vdW dE dE_avg Temp dG
> # l l+dl l l+dl E(l+dl)-E(l)
> #NEW FEP WINDOW: LAMBDA SET TO 1 LAMBDA2 0.98
> FepEnergy: 10 -2364.7855 -2364.6734 -1663.4688 -1663.4864 0.0946 0.2396 194.4961 0.2345

And here's what I get on this computer shortly before it crashes:

> # STEP Elec vdW dE dE_avg Temp dG
> # l l+dl l l+dl E(l+dl)-E(l)
> #NEW FEP WINDOW: LAMBDA SET TO 1 LAMBDA2 0.98
> FepEnergy: 10 -226.4117 12919.7280 -2121.0779 -2121.2665 13145.9511 13076.4999 142.9543 9999999999.9999

It seems the electrostatic energy at the new value of lambda is not calculated correctly. This is NAMD 2.9, using FFTW 3.3. The version that worked was using FFTW 2.1.3. Any ideas? Has anyone seen this before?

Thanks!
JC

• Next message: Vidhya Sankar: "About Muticore Linux Ibnstalltion if NAmd"
• Previous message: Francesco Pietra: "Re: About the scope of show_replicas.vmd"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:23:45 CST