[MITgcm-support] Fwd: Scalability on a new Sgi node

[Stefano Querin]

Second part.

Thanks,

S.


Begin forwarded message:

> From: Stefano Querin <squerin at ogs.trieste.it>
> Date: 06 July 2011 18:29:07 GMT+02:00
> To: Jean-Michel Campin <jmc at ocean.mit.edu>
> Subject: Fwd: [MITgcm-support] Scalability on a new Sgi node
>
> Hi Jean-Michel,
>
> for the sake of completeness, I'm attaching the results of the last  
> simulation (on the stand alone node, Ubuntu OS and up to date PGI  
> compiler - 11.6).
> If you look at the final statistics, the wall clock time is  
> abnormal...
>
> User time:  1323.599985351786
> System time:  716.6300000026822
> Wall clock time:  15028.45807790756
>
> These are the results running on the same node, when mounted on the  
> cluster (with Rocks OS and old version of PGI - 6.1):
>
> 24 (6x4) cobra-0-5
> User time: 1704.439960937947
> System time: 124.3899948149920
> Wall clock time: 2270.645053148270
>
> and these are the results on the other HPC cluster:
>
> 24 (6x4)
> User time: 328.649038668722
> System time: 0.000000000000000E+000
> Wall clock time: 360.033116102219
>
> There must be something VERY wrong in some configuration/option...
>
> Cheers!
>
> SQ
>
>
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: STDOUT.0000
Type: application/octet-stream
Size: 943078 bytes
Desc: not available
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20110721/22a9cf72/attachment-0001.obj>
-------------- next part --------------

>