Hello - again
Replying to my own mail.
I had version 3.6.3 installed as well - linking to this version I get my
100% on one core back.
NetCDF4 I've compiled against both 1.8.6 and 1.8.8 of hdf - and in both
cases I only get around 50% CPU utilisation.
Karsten
On Mon, Dec 12, 2011 at 07:37:36 +0000, Karsten Bolding wrote:
> Hello
>
> I use NetCDF on a number of different Linux machines. My personal
> workstation I normally keep up to date with the newest version of NetCDF
> (4.1.3 and 4.2 for Fortran).
>
> A standard test case I'm using shows a CPU-utilisation of about 50% on
> one core - if I switch off saving during the simulation I get my 100%
> back.
>
> Now on a different machine I can get my - very close to - 100% also in
> the case when I do saving. The size of the output file is only about
> 55MB and the model runs is like 30 seconds on the machine with full CPU
> utilisation.
>
> I know it might no be the fault of NetCDF. On my work station I've
> Ubuntu 11.10 installed - but on the production machine I've Debian Sid
> (not updated for a long time).
>
> Anybody else experiencing problems with NetCDF performance - or shall I
> blame it on Ubuntu :-)
>
> Karsten
>
>
>
>
> --
> ----------------------------------------------------------------------
> Karsten Bolding Bolding & Burchard ApS
> Strandgyden 25 Phone: +45 64422058
> DK-5466 Asperup Fax: +45 64422068
> Denmark Email: karsten@xxxxxxxxxxxxxxxxxxxx
>
> http://www.findvej.dk/Strandgyden25,5466,11,3
>
> Q: What is irony?
> A: Almost the same as goldy and bronzy - just made of iron ...
> (from Blackadder)
> ----------------------------------------------------------------------
>
--
----------------------------------------------------------------------
Karsten Bolding Bolding & Burchard ApS
Strandgyden 25 Phone: +45 64422058
DK-5466 Asperup Fax: +45 64422068
Denmark Email: karsten@xxxxxxxxxxxxxxxxxxxx
http://www.findvej.dk/Strandgyden25,5466,11,3
Q: What is irony?
A: Almost the same as goldy and bronzy - just made of iron ...
(from Blackadder)
----------------------------------------------------------------------
