climateprediction.net home page
Iceworlds & Slowdowns hadsm3/mh - Closed - Discussion

Iceworlds & Slowdowns hadsm3/mh - Closed - Discussion

Message boards : Number crunching : Iceworlds & Slowdowns hadsm3/mh - Closed - Discussion
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 . . . 15 · Next

AuthorMessage
Tabaluga

Send message
Joined: 27 Aug 04
Posts: 3
Credit: 1,700,141
RAC: 0
Message 31896 - Posted: 27 Dec 2007, 10:12:59 UTC

another slow iceworld

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=6790782

timestamp 57730

4,36 sec/TS

global is blue

Intel pentium M 1700MHz, never overclocked

Should i abort this model ?
I there any helpful outcome ?




ID: 31896 · Report as offensive
Profile Iain Inglis

Send message
Joined: 9 Jan 07
Posts: 467
Credit: 14,450,296
RAC: 0
Message 31897 - Posted: 27 Dec 2007, 11:17:12 UTC
Last modified: 27 Dec 2007, 11:20:08 UTC

The model (here) seems to have slowed down by a factor of more than ten and is only five timesteps into phase 3. Nothing will be reported for that phase until the end, so the model has another 19 steps to go before any useful new information is sent to the server - which will take forever.

Abort it.

(And well done for searching the message boards!)

[Edit: oops, Les has answered in another thread.]
ID: 31897 · Report as offensive
Profile JIM

Send message
Joined: 31 Dec 07
Posts: 1092
Credit: 19,659,569
RAC: 5,667
Message 32305 - Posted: 23 Jan 2008, 16:54:46 UTC

Hi, everyone.

I am running the hadsm3 model and it was progressing normally until 2 days ago. At about 93% the model graphic for temp. went uniformly blue. Temp scale shows -42 C worldwide. I think the model has gone ice world. The cloud graphic is also abnormal. All clouds are gone and oceans are bluish white. Other than that it seems to be progressing at normal speed. I restored from a backup and it ran normally for 2 days and then went ice world again at the same point.

I plan to run it to the end because time to completion only 29 hours. I hope the results will be useful to the guys at Oxford for something.

ID: 32305 · Report as offensive
Profile MikeMarsUK
Volunteer moderator
Avatar

Send message
Joined: 13 Jan 06
Posts: 1498
Credit: 15,613,038
RAC: 0
Message 32308 - Posted: 23 Jan 2008, 18:10:20 UTC


As long as it is running at normal speed it is best to let it run - the scientists want to know which combinations of parameters lead to viable climates, and which don\'t.
I'm a volunteer and my views are my own.
News and Announcements and FAQ
ID: 32308 · Report as offensive
Profile JIM

Send message
Joined: 31 Dec 07
Posts: 1092
Credit: 19,659,569
RAC: 5,667
Message 32322 - Posted: 24 Jan 2008, 5:05:09 UTC - in response to Message 32308.  


As long as it is running at normal speed it is best to let it run - the scientists want to know which combinations of parameters lead to viable climates, and which don\'t.


Thanks, I was going to let it run anyway because there is less than one day left on the \"to completion.\" timer. Happy to hear that an ice world result will be useful to the climete scientists.
ID: 32322 · Report as offensive
Profile John Hunt
Avatar

Send message
Joined: 5 Mar 05
Posts: 64
Credit: 790,577
RAC: 0
Message 32526 - Posted: 8 Feb 2008, 7:56:09 UTC


Another slow iceworld.

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7209853

Timestep 241564 of 259248 - 97.73% complete. ( 0.255% more than 24 hours ago)

CPU time = 296:31 hrs, 6:53 to completion.

1.40 sec/TS.

Globe is blue.

Intel Q6600 Quad, never overclocked.

Should I keep going?

Model date is 23/11/2064 so it looks as if I\'m going to have a verrrry chilly 109th birthday on the 24th!
ID: 32526 · Report as offensive
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7076
Credit: 21,440,546
RAC: 15,333
Message 32527 - Posted: 8 Feb 2008, 8:06:19 UTC


I had one on a quad, which was a \'slow processor\', and about one trickle from the finish. I let it run until the other 3 finished, and then aborted it. It never did get to the next trickle.

A \'fast processor\' may make it to the end.

ID: 32527 · Report as offensive
Profile John Hunt
Avatar

Send message
Joined: 5 Mar 05
Posts: 64
Credit: 790,577
RAC: 0
Message 32528 - Posted: 8 Feb 2008, 8:24:16 UTC


Thanks Les!
Apart from myself, there\'s two duo cores and a quad core (and a Celeron!) also crunching this WU.
I\'ll give it a few more days, see what happens. If and when the other quad/duo results come in, I\'ll abort.


ID: 32528 · Report as offensive
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 13,458,077
RAC: 5,270
Message 32529 - Posted: 8 Feb 2008, 14:32:39 UTC

John, just to compare your iceworld graphics with a beta model I ran that went seriously wrong, could you tell me what sort of display you see when you press R on the keyboard for rain & snow, and P for pressure? Do you see a changing and progressing display, or just a single uniform colour?
Cpdn news
ID: 32529 · Report as offensive
Profile John Hunt
Avatar

Send message
Joined: 5 Mar 05
Posts: 64
Credit: 790,577
RAC: 0
Message 32532 - Posted: 8 Feb 2008, 17:17:42 UTC
Last modified: 8 Feb 2008, 17:25:50 UTC

Sorry for delay in reply -

R for rainfall gives a static pastel green globe.
P for pressure gives a static black globe.
Both have white outlines for land.

As I said to Les earlier, I don\'t normally watch the graphics.....

[EDIT]
Computer 827039 may have hit the same problem as myself - no trickle since early hours of this morning....
ID: 32532 · Report as offensive
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 13,458,077
RAC: 5,270
Message 32535 - Posted: 8 Feb 2008, 17:53:11 UTC

Thanks, John. If your model does make it to completion I\'ll be interested to see its graphs.
Cpdn news
ID: 32535 · Report as offensive
Profile John Hunt
Avatar

Send message
Joined: 5 Mar 05
Posts: 64
Credit: 790,577
RAC: 0
Message 32542 - Posted: 9 Feb 2008, 6:19:55 UTC - in response to Message 32535.  

Thanks, John. If your model does make it to completion I\'ll be interested to see its graphs.


Almost 24 hours on from first post yesterday and it\'s now 97.22% complete.
At this rate, it\'ll be 10 days + to completion.
I\'ll see this one through!

ID: 32542 · Report as offensive
Profile JIM

Send message
Joined: 31 Dec 07
Posts: 1092
Credit: 19,659,569
RAC: 5,667
Message 32555 - Posted: 10 Feb 2008, 7:45:41 UTC - in response to Message 32542.  

Thanks, John. If your model does make it to completion I\'ll be interested to see its graphs.


Almost 24 hours on from first post yesterday and it\'s now 97.22% complete.
At this rate, it\'ll be 10 days + to completion.
I\'ll see this one through!



You have my condolences on your model going iceball. I know how you feel. I recently had an HadSM model do the same thing at about 93%. I restored it from a backup, but, it went iceball again at the same point so I knew it was a flaw in the model and not some problem with my machine. Since it seemed to be progressing at normal speed I let it run to the end.

Even these semi-failed models produce some useful data. They help the climate scientists learn which starting parameters yield viable climates, and which do not. So keep on crunching.

ID: 32555 · Report as offensive
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 13,458,077
RAC: 5,270
Message 32568 - Posted: 10 Feb 2008, 22:14:51 UTC
Last modified: 10 Feb 2008, 22:27:18 UTC

Hi Jim

I think I\'ve found it - this one

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?field=Precipitation&resultid=7080467&phase=AP#graph
which did speed up at the end and where the precipitation towards the end of phase 3 stopped being processed.

What did you notice in the graphics while it was still crunching? Did it give an all-blue graph for temperatures?
Cpdn news
ID: 32568 · Report as offensive
Profile JIM

Send message
Joined: 31 Dec 07
Posts: 1092
Credit: 19,659,569
RAC: 5,667
Message 32570 - Posted: 11 Feb 2008, 1:17:11 UTC - in response to Message 32568.  

Hi Jim

I think I\'ve found it - this one

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?field=Precipitation&resultid=7080467&phase=AP#graph
which did speed up at the end and where the precipitation towards the end of phase 3 stopped being processed.

What did you notice in the graphics while it was still crunching? Did it give an all-blue graph for temperatures?


Hi, this is Jim.

Yes the temp. graphic was uniformly blue with the temp. scale indicating -42 C or colder all the way from the poles to the equator.
The cloud graphic is also abnormal. All clouds are gone and oceans are white.

ID: 32570 · Report as offensive
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 13,458,077
RAC: 5,270
Message 32571 - Posted: 11 Feb 2008, 2:07:26 UTC

Did you see the temperatures gradually turning bluer and bluer or was everything suddenly completely blue one day when you looked?

I now think that an all-blue globe doesn\'t necessarily mean low temperatures. I think it\'s just as likely to indicate that the calculations have failed in some way. The uniform blue must be the default colour indicating no data input to generate real graphics. Your phase 3 temp graph doesn\'t indicate freezing temperatures world-wide, even at the end.

It\'s probably just coincidence that the default colour is the same colour as for low temps. If the default colour was red we\'d be saying we have furnace or oven worlds.

Can you remember whether P for atmospheric pressure also produced a single uniform colour?
Cpdn news
ID: 32571 · Report as offensive
Profile JIM

Send message
Joined: 31 Dec 07
Posts: 1092
Credit: 19,659,569
RAC: 5,667
Message 32574 - Posted: 11 Feb 2008, 6:15:01 UTC - in response to Message 32571.  

Did you see the temperatures gradually turning bluer and bluer or was everything suddenly completely blue one day when you looked?

I now think that an all-blue globe doesn\'t necessarily mean low temperatures. I think it\'s just as likely to indicate that the calculations have failed in some way. The uniform blue must be the default colour indicating no data input to generate real graphics. Your phase 3 temp graph doesn\'t indicate freezing temperatures world-wide, even at the end.

It\'s probably just coincidence that the default colour is the same colour as for low temps. If the default colour was red we\'d be saying we have furnace or oven worlds.

Can you remember whether P for atmospheric pressure also produced a single uniform colour?


Hi, this is Jim

The color went from normal to all blue suddenly. The temp. graphic looked normal when I checked on it before I went to bed and the following morning (I let it run all night) it had gone all blue.

Sorry, I didn’t check the pressure.

ID: 32574 · Report as offensive
Profile John Hunt
Avatar

Send message
Joined: 5 Mar 05
Posts: 64
Credit: 790,577
RAC: 0
Message 32590 - Posted: 12 Feb 2008, 19:22:37 UTC
Last modified: 12 Feb 2008, 19:28:55 UTC

I\'m still hanging in there with this one!
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7209853

6 days from previous trickle to the one today - just one more trickle to go. (phew!)
ID: 32590 · Report as offensive
Profile John Hunt
Avatar

Send message
Joined: 5 Mar 05
Posts: 64
Credit: 790,577
RAC: 0
Message 32670 - Posted: 19 Feb 2008, 20:46:49 UTC - in response to Message 32590.  

I\'m still hanging in there with this one!
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7209853

6 days from previous trickle to the one today - just one more trickle to go. (phew!)


WU finished and returned at lunchtime UTC today 19/02/2008.
Hope the results are useful to the scientists!


ID: 32670 · Report as offensive
Profile MikeMarsUK
Volunteer moderator
Avatar

Send message
Joined: 13 Jan 06
Posts: 1498
Credit: 15,613,038
RAC: 0
Message 32673 - Posted: 19 Feb 2008, 21:28:09 UTC


Congratulations :-)

I'm a volunteer and my views are my own.
News and Announcements and FAQ
ID: 32673 · Report as offensive
Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 . . . 15 · Next

Message boards : Number crunching : Iceworlds & Slowdowns hadsm3/mh - Closed - Discussion

©2019 climateprediction.net