Re: AMBER: AMBER8 compiler with intel 9.1 compiler

From: Peter Z. Qin <pzq.usc.edu>
Date: Fri, 25 May 2007 09:49:13 -0700

I just found out there is one more failure in the test using AMBER8 compiled
with intel 9.1 but with FOPTFLAGS set to the same as FFLAGS.

[pzq.hpc-login1 pbsa_dmp]$ more mdout.dmp.min.dif
62c62
< 1 5.9005E+2 2.4154E+2 8.2881E+2 C2 1 
---
>       1      -7.0318E+6     2.4154E+2     8.2881E+2     C2          1
65a66
>  EPOLAR  =        0.0000  ETHREEB = -7032360.0000
69c70
<       1       5.9005E+2     2.4154E+2     8.2881E+2     C2          1
---
>       1      -7.0318E+6     2.4154E+2     8.2881E+2     C2          1
72a74
>  EPOLAR  =        0.0000  ETHREEB = -7032360.0000
The difference between in ENERGY (5.9E2 vs. -7.0E6) seems huge to me.  And 
why would the new file output extra EPOLAR = 0?
We still have not figured out what the problem is with the intel 9.1 
optimizer.  Will this help, or is this a different bug?
-peter
----- Original Message ----- 
From: "David A. Case" <case.scripps.edu>
To: <amber.scripps.edu>
Sent: Thursday, May 17, 2007 6:38 PM
Subject: Re: AMBER: AMBER8 compiler with intel 9.1 compiler
>>
>> The tests of serail version all PASSED (except for the antechamber, which 
>> I
>> will post in a separate email).  Tests of parallel version of snader and
>> sander.LES returned only one possible failure:
>>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sun May 27 2007 - 06:07:46 PDT
Custom Search