AMBER Archive (2007)

Subject: Re: AMBER: AMBER8 compiler with intel 9.1 compiler

From: Peter Z. Qin (pzq_at_usc.edu)
Date: Fri May 18 2007 - 19:53:29 CDT


----- Original Message -----
From: "David A. Case" <case_at_scripps.edu>
To: <amber_at_scripps.edu>
Sent: Thursday, May 17, 2007 6:38 PM
Subject: Re: AMBER: AMBER8 compiler with intel 9.1 compiler

> 2. If so, try just compiling ew_fft.f (and/or ew_recip.f) at low
> optimization, everything else at high optimzation. My guess is that
> ew_fft.f
> is the culprit...there seem to be things in there that the intel
> optimizers
> don't like.
>
I modified the ew_fft.f by removing "! <compile=optimized>" and also applied
budfix64. Then I recompile with optimization on. However, failures occur at
the same cases (cytosine, dhfr, etc.)

>
> If you want, the next thing to do is to try to figure out which routine
> doesn't
> work with optimization turned on. Here is a suggested path, (for you or
> someone else):
>
> 1. Do dhfr one step with verbose=2 in the &ewald namelist. Look at the
> differences between optimized and unoptimzed code. If you are lucky, the
> differences will be in the reciprocal contribution to the electrostatic
> energy.

With verbose=2, I have ran dhfr/Run.dhfr. Part of the difference between
mdout.dhfr is listed below. The version with MPI flag is compiled without
optimization, and the version with no Flags is compiled as indicated above.

The big difference seems to be Eevir vs. Ecoulomb.

188,189c191,192
< Iso virial = -187936.116173447459
< Eevir vs. Ecoulomb = 0.000535006138

---
>      Iso virial             =     -187933.775401700172
>      Eevir vs. Ecoulomb     =           0.000037834768

What next?

-peter

[pzq_at_hpc-login1 dhfr]$ more dif.op_nonop 6c6 < | Run on 05/18/2007 at 17:25:32 --- > | Run on 05/18/2007 at 17:31:37 43c43 < | Flags: --- > | Flags: MPI 80c80 < | Integer memory used by local nonb= 1113175 --- > | Integer memory used by local nonb= 855625 83c83 < | MAX NONBOND PAIRS = 7643333 --- > | MAX NONBOND PAIRS = 3821666 86c86 < | Real 1587417 --- > | Real 1656247 88,90c88,90 < | Integer 1822718 < | Max Pairs 7643333 < | Max Rstack 1147486 --- > | Integer 1565168 > | Max Pairs 3821666 > | Max Rstack 1017678 92c92 < | Total 59355 kbytes --- > | Total 42944 kbytes 151a152,155 > | Atom division among processors: > | 0 11467 22930 > | Running AMBER/MPI version on 2 nodes > 178,179c182 < listtot = 4721242 < | Local SIZE OF NONBOND LIST = 4721242 --- > | Local SIZE OF NONBOND LIST = 2351245 183c186 < Evdw = 8109.389249773295 --- > Evdw = 8109.389249773089 185c188 < Ecoulomb = -89881.644058586680 --- > Ecoulomb = -89881.644058587204 188,189c191,192 < Iso virial = -187936.116173447459 < Eevir vs. Ecoulomb = 0.000535006138 --- > Iso virial = -187933.775401700172 > Eevir vs. Ecoulomb = 0.000037834768 192,194c195,197 < (rec) = 791.295774837328 < (dir) = -70279.214671935872 < (adj) = 463442.359416397696 --- > (rec) = 791.295774837370 > (dir) = -70279.214671936992 > (adj) = 463442.359416398278 196,198c199,201 < MOLECULAR VIRIAL: -.61556609E+05 -.73132351E+03 0.16272816E+03 < MOLECULAR VIRIAL: -.73132351E+03 -.63437371E+05 0.39174475E+03 < MOLECULAR VIRIAL: 0.16272816E+03 0.39174475E+03 -.62942136E+05 --- > MOLECULAR VIRIAL: -.61554269E+05 -.73132351E+03 0.16272816E+03 > MOLECULAR VIRIAL: -.73132351E+03 -.63460369E+05 0.39174475E+03 > MOLECULAR VIRIAL: 0.16272816E+03 0.39174475E+03 -.62919138E+05 200,201c203,204 < NSTEP = 1 TIME(PS) = 510.051 TEMP(K) = 301.30 PRESS = 0.0 < Etot = -57751.4656 EKtot = 14110.8966 EPtot = -71862.3622 --- > NSTEP = 1 TIME(PS) = 510.051 TEMP(K) = 301.29 PRESS = > 0.0 > Etot = -57751.8286 EKtot = 14110.5335 EPtot > = -71862.3622 205c208 < Ewald error estimate: 0.5350E-03 --- > Ewald error estimate: 0.3783E-04 208c211 < NET FORCE PER ATOM: -0.2246E-04 0.1758E-04 -0.4452E-05 --- > NET FORCE PER ATOM: -0.2246E-04 0.1757E-04 -0.4451E-05 210c213 < Evdw = 8112.296676774401 --- > Evdw = 8112.287287794546 212c215 < Ecoulomb = -89882.083206685784 --- > Ecoulomb = -89882.066896236734 215,216c218,219 < Iso virial = -187985.022914657660 < Eevir vs. Ecoulomb = 0.000546464804 --- > Iso virial = -187983.277227977349 > Eevir vs. Ecoulomb = 0.000026650466 219,221c222,224 < (rec) = 789.943303354934 < (dir) = -70279.067483709237 < (adj) = 463443.125551554316 --- > (rec) = 789.940498351458 > (dir) = -70279.064612572838 > (adj) = 463443.141795870382 223,225c226,228 < MOLECULAR VIRIAL: -.61574500E+05 -.71646918E+03 0.18477125E+03 < MOLECULAR VIRIAL: -.71646918E+03 -.63433957E+05 0.36468806E+03 < MOLECULAR VIRIAL: 0.18477125E+03 0.36468806E+03 -.62976566E+05 --- > MOLECULAR VIRIAL: -.61573011E+05 -.71651922E+03 0.18472266E+03 > MOLECULAR VIRIAL: -.71651922E+03 -.63455503E+05 0.36481257E+03 > MOLECULAR VIRIAL: 0.18472266E+03 0.36481257E+03 -.62954763E+05

----------------------------------------------------------------------- The AMBER Mail Reflector To post, send mail to amber_at_scripps.edu To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu