Hi all.
I just performed a compilation of ambertools23+amber22 with gcc7.5 and
intel mkl (latest version, it was installed yesterday... :P ).
The ambertools serial compilation went smoothly {2886 file comparisons
passed, 4 file comparisons failed (4 of which can be ignored), 0 tests
experienced errors}, while amber had some bumps in the road... Amber
serial summary of tests:
###########
possible FAILURE: check mdout.trxox.dif
possible FAILURE: check mdout.trxox_md.dif
possible FAILURE: check mdout.trxox_md.dif
possible FAILURE: (ignored) check mdout.mcwat.dif
possible FAILURE: check mdout.1hji_md.dif
possible FAILURE: check mdout.3p4a.dif
possible FAILURE: (ignored) check mdout.dif
###########
Whose numbers reported are severely different, despite being just a few
of them (as the attachment shows, there are strange gbalphaP, gbbetaP
and gbgammaP with completely different numbers than the expected)as can
be seen in the 2023-11-08_02-33-56.diff attached file.
In parallel I'm however having more issues...
Summary of AmberTools23 parallel tests:
###########
1314 file comparisons passed
7 file comparisons failed (0 of which can be ignored)
8 tests experienced errors
Test log file saved as
/usr/local/chem/amber23///logs/test_at_parallel/2023-11-08_16-05-06.log
Test diffs file saved as
/usr/local/chem/amber23///logs/test_at_parallel/2023-11-08_16-05-06.diff
###########
Greping the log file, I got that:
###########
possible FAILURE: check bbmd.1.txt.dif
possible FAILURE: check mdout.1.dif
possible FAILURE: check bbmd.2.txt.dif
possible FAILURE: check bbmd.3.txt.dif
possible FAILURE: check umbrella.1.ncdump.dif
possible FAILURE: check umbrella.2.ncdump.dif
possible FAILURE: check umbrella.3.ncdump.dif
###########
Looking at the .dif file, bbmd.1, mdout.1 and umbrella.1 have just small
differences in the last digit, however files .2 and .3 indicate much
higher differences (see 2023-11-08_15--5-06.diff attatched file)... :(
Moreover, the summary of Amber22 parallel tests:
############
167 file comparisons passed
10 file comparisons failed (2 of which can be ignored)
0 tests experienced an error
Test log file saved as
/usr/local/chem/amber23/logs/test_amber_parallel/2023-11-08_16-36-08.log
Test diffs file saved as
/usr/local/chem/amber23/logs/test_amber_parallel/2023-11-08_16-36-08.diff
############
And greping the log file I get:
############
possible FAILURE: check mdout.trxox.dif
possible FAILURE: check mdout.trxox_md.dif
possible FAILURE: check mdout.trxox_md.dif
possible FAILURE: check mdout.1hji_md.dif
possible FAILURE: (ignored) check mdout.dif
possible FAILURE: (ignored) check campTI.out.dif
possible FAILURE: check rep-1.mdout.dif
possible FAILURE: check rep-1.lam.dif
possible FAILURE: check rem.log.dif
possible FAILURE: check mdout.MPI.dif
############
Looking in the 2023-11-08_16-36-08.dif file, now all of them seems
pretty severe to me... :(
I'm using the latest version of mpich compiled specifically for
amber/ambertools as suggested in the manuals, by the way.
Does anybody has any idea o the possible cause for this error?
Thanks a lot in advance for any help!
Best regards,
Jones
--
Jones de Andrade
(jdandrade.iq.ufrgs.br)
DFQ/IQ/UFRGS
Lattes: http://lattes.cnpq.br/6675936210583999
Orcid: https://orcid.org/0000-0003-3429-8119
ResearcherID: https://publons.com/researcher/AAC-5337-2019/
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Nov 08 2023 - 19:00:02 PST
