[MITgcm-support] MITgcm on AS4, kernel 2.6

[Hezi Gildor]

Hi Ed,

thanks for the prompt reply.

The OS was AS4.0 and AS4U1. The compilers tried are gcc and pgi (5.2 and 
6.0). Tried using with LAM, MPICH (from pgi as well as built from tar 
ball), MPICH2. The compiler options used were many. From the default 
options to the -O3, to -fast, -fastsse, and -kieee (last 3 for pgi).

Communications is through the internet. (No private network for now).

Dual-opteron machines with

AMD Opteron(tm) Processor 250

The code is based on MITgcm_ss_20050719, checkpoint571_post. We have few 
hard-wired changes to code but it runs ok on Alpha machine and on 
cluster with Linux Kernel 2.4 so i don't think that the problem is with 
the code.

when we compared fields such as salinity between two runs with same 
executables, we can get that at most grid points the difference is zero 
but at many isolated points scattered throughout the domain it is not (O 
of 10^-6).

thanks, hezi.


Ed Hill wrote:

>On Thu, 2005-10-06 at 16:50 +0300, Hezi Gildor wrote:
>  
>
>>Dear all,
>>
>>did anyone used the MITgcm on Opteron cluster running AS4 or 4.1 (Linux 
>>Kernel 2.6)?
>>
>>we tried on on a local machine and encounter strange behavior. e.g., 
>>running the same executable twice gives somewhat different results. On 
>>Linux kernel 2.4 it seems to work ok.
>>    
>>
>
>Hi Hezi,
>
>We've run MITgcm on a number of different Opteron systems (some of which
>are listed on the testing pages at: http://mitgcm.org/testing.html ) and
>have yet to experience the problem that you're describing.  And we use
>Fedora and Red Hat (RHEL) on the majority of our systems so we're very
>confident that MITgcm works nicely on them.
>
>Is there any chance that you could be a little more specific about both
>your setup (OS, compiler, etc.) and, more importantly, the sort of
>error(s) you're seeing?
>
>Ed
>
>  
>