[MITgcm-support] Poor cpu usage percentage

[Dimitris Menemenlis]

Wild guesses again:

(1) If your macbook has solid state disk vs hard disk on your iMac, it could still be an IO problem.
Look at tail end of STDOUT.0000 to see the timings and find out which component of MITgcm
are slower on iMac vs macbook.

(2) Your compilation options or compilers differ.  Look at your Makefile in build directory to see
whether you are using same compiler and optimization options.

> On Jul 20, 2015, at 8:12 AM, Rogan, Nicholas <nrogan at geomar.de> wrote:
> 
> Hi Dimitris,
> Thanks for getting back to me. Thanks for correcting me, I was under the impression that make -j 2 was to do with the runtime use of multiple cores. 
> 
> So, I have been running my MITgcm setup on  my macbook pro for the last fw months. I had problems with my iMac and I could not ssh in to carry out the runs on there (It is currently sitting in a country other than the one I'm working). I was back visiting last week and got it up and running again. I am purely using it for running my model. 
> For comparison:
> My iMac is        2.66 GHz Intel core 2 duo with 8Gb RAM.
> My macbook is   2.4  Ghz Intel core 2 duo with 4Gb RAM.
> 
> I compiled the MITgcm identically on both machines, but currently my Macbook is running about twice as fast as my iMac. The macbook usually has 5-10% IDLE CPU usage whereas the iMac is running with 65-90% IDLE CPU usage.They are both run with identical data files and the only program running on both (other than hidden processes) is Terminal. 
> 
> So I'm pretty sure the issue is not the I/O as I'm comparing relative to my macbook and I have no other programs running. I've attached screenshot of the top details. It's very strange, I've never had this problem before.
> 
> Nick

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20150720/2118ad94/attachment.htm>