On Mar 28, 2011, at 8:58 AM, Ben Roberts wrote:
> Hi all,
>
> On 27/3/2011, at 7:39 a.m., case wrote:
>
>> Hi everyone:
>>
>> I have posted a new release candidate here:
>>
>> http://ambermd.org/downloads/AmberTools.26mar11.tar.bz2
>>
>> Still needed:
>>
>> 1. Proofreading of Users' Manual.
>> 2. Testing of the ff10 force field files
>> 3. Minor tweaks to the tests, esp. MMPBSA.py
>
> For what it's worth, using Intel 11.1.089 and MKL on Mac OS X 10.6.6, I get five tests failing with numerical differences between the output of the program and the reference file:
>
> nab/asp.out
> nab/aspnb.out
> nab/asp_nm1.out
> mmpbsa_py/06_NAB_Nmode/FINAL_RESULTS_MMPBSA.dat
> sqm/ch3f/ch3f.pm6.go.out
The sqm failure is OK. In geometry optimizations it typically is only the orientation of the molecule that differs on different platforms due to small numerical differences.
> I've attached the full diff. Can anyone clarify for me whether the results I got are within a tolerable margin?
>
> I hope to take a peek at the users' manual shortly, too.
>
> Cheers,
> Ben
> <2011-03-28_11-10-14.diff>
> --
> For greater security, I support S/MIME encryption.
>
> _______________________________________________
> AMBER-Developers mailing list
> AMBER-Developers.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber-developers
--
Dr. Andreas W. Goetz
San Diego Supercomputer Center
Tel : +1-858-822-4771
Email: agoetz.sdsc.edu
Web : www.awgoetz.de
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Mon Mar 28 2011 - 13:30:05 PDT
