case wrote:
> On Wed, Mar 10, 2010, Mark Williamson wrote:
>
>>>> test/gb_rna
>>> I don't see any errors in gb_rna at either the "-np 2" or "-np 4" levels.
>>> Do you mean you get these errors with your code, or with what is in CVS?
>>> If the latter, can you give compiler details?
>>>
>> Sorry for being unclear here; I am seeing these with the current amber11
>> CVS tree. Here are the requested details:
>>
>> [bunny:0.00][mjw:test]$ ifort -V
>> Intel(R) Fortran Compiler for applications running on Intel(R) 64,
>> Version 10.1 Build 20080801 Package ID: l_fc_p_10.1.018
>> Copyright (C) 1985-2008 Intel Corporation. All rights reserved.
>> FOR NON-COMMERCIAL USE ONLY
>
> Weird. If I have a chance, I'll take a look with ifort 10.1.018. But what
> *are* the errors (i.e. clearly more than roundoff?)
>
> ...thx...dave
I just want to follow up on specifically the following failures:
test/ti_decomp/ti_decomp_1.out.dif
test/ti_decomp/ti_sc_decomp_1.out.dif
test/ti_decomp/ti_sc_decomp_2.out.dif
I do not believe this is a compiler bug, since I am seeing it with two
different compilers (ifort 10.1.018 and gfortran 4.4.3, both with
mpich2-1.0.7) and the error is the same:
http://www.wmd-lab.org/mjw/for_amber_dev/180310/
Is there something up with the decomp code itself?
regards,
Mark
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Thu Mar 18 2010 - 17:00:03 PDT
