climateprediction.net home page
Speed variation between models

Speed variation between models

Message boards : Number crunching : Speed variation between models
Message board moderation

To post messages, you must log in.

AuthorMessage
old_user45757

Send message
Joined: 29 Jan 05
Posts: 4
Credit: 120,918
RAC: 0
Message 8491 - Posted: 2 Feb 2005, 21:24:51 UTC

Hi!
I have two different climate models running on my computer (accidentaly I attached to a project a second time), each taking up about 50% of the processor power (when nothing else is being used). One averages 3.37 s/Ts the other 3.44 s/Ts. Is this difference due to the way my processor assigns processor power to the first process preferentially, or do the varation of the parameters within the models cause the difference in speed?
Just curious...
ID: 8491 · Report as offensive     Reply Quote
Profile geophi
Volunteer moderator

Send message
Joined: 7 Aug 04
Posts: 2169
Credit: 64,555,079
RAC: 5,903
Message 8492 - Posted: 2 Feb 2005, 21:35:14 UTC - in response to Message 8491.  

> Hi!
> I have two different climate models running on my computer (accidentaly I
> attached to a project a second time), each taking up about 50% of the
> processor power (when nothing else is being used). One averages 3.37 s/Ts the
> other 3.44 s/Ts. Is this difference due to the way my processor assigns
> processor power to the first process preferentially, or do the varation of the
> parameters within the models cause the difference in speed?
> Just curious...
>

It's the 2nd one. Different parameters result in different calculation times.
ID: 8492 · Report as offensive     Reply Quote
Profile old_user36084
Avatar

Send message
Joined: 15 Jan 05
Posts: 31
Credit: 1,249,348
RAC: 0
Message 8495 - Posted: 2 Feb 2005, 22:47:17 UTC

On my two HT CPUs slot 0 is always faster than slot 1, on my limited comparison of running my first 6 models.
ID: 8495 · Report as offensive     Reply Quote
old_user23880
Volunteer tester

Send message
Joined: 10 Oct 04
Posts: 223
Credit: 4,664
RAC: 0
Message 8500 - Posted: 3 Feb 2005, 1:08:41 UTC

If you watched the picture globe of one model preferentially, or left one on accidentally overnight, that would be enough to make a difference.
__________________________________________________

ID: 8500 · Report as offensive     Reply Quote

Message boards : Number crunching : Speed variation between models

©2024 climateprediction.net