climateprediction.net home page
Time to Completion and Sulphur Model

Time to Completion and Sulphur Model

Questions and Answers : Preferences : Time to Completion and Sulphur Model
Message board moderation

To post messages, you must log in.

AuthorMessage
old_user162591

Send message
Joined: 15 Feb 06
Posts: 3
Credit: 5,568
RAC: 0
Message 23192 - Posted: 18 Jun 2006, 6:24:41 UTC

I\'ve clocked up 665 hours (27.50%)under sulphur model sulphur_i36h_200843929_0 in application sulphur_cycle 4.22. Yesterday, there were 2380 hours to completion.
I have two queries:
1. Why, this morning, with the program suspended overnight, has the time to completion increased to 2,648 hours (percentage completed remains at 27.50)?
2. By chance, I\'ve read on your message board that the sulphur model is obsolete, so am I wasting my time and is there no way that this fact could have been communicated direct to me? Should I now delete the project?
Thanks.
ID: 23192 · Report as offensive     Reply Quote
Profile MikeMarsUK
Volunteer moderator
Avatar

Send message
Joined: 13 Jan 06
Posts: 1498
Credit: 15,613,038
RAC: 0
Message 23194 - Posted: 18 Jun 2006, 8:45:59 UTC

Hi Chris,

The \'estimated time to completion\' is usually very misleading, particularly with sulphur, and I usually look at the % complete combined with the CPU hours. Using those two, I calculate that you havve 1753 CPU hours remaining.

As for the \'obsolete\' question, I can\'t give you any indications because I don\'t know, but be aware that the coupled model is much larger than the sulphur model, and would probably take something like 7000 CPU hours to complete. So it\'s probably best to stick with Sulphur for the time being.

There is a shorter (but more memory-hungry) climate model, the \'seasonal attribution model\'. This takes about half the time to run as the Sulphur model, and a sixth of the time to run as the coupled model, but does need more memory. Take a look at http://seasonal.cpdn.org/ if you\'re interested in that.
I'm a volunteer and my views are my own.
News and Announcements and FAQ
ID: 23194 · Report as offensive     Reply Quote
old_user162591

Send message
Joined: 15 Feb 06
Posts: 3
Credit: 5,568
RAC: 0
Message 23195 - Posted: 18 Jun 2006, 11:48:50 UTC - in response to Message 23194.  

Thanks, Mike. That\'s very helpful on the first issue, but I\'m not inclined to let my PC whirr away for months on a project that won\'t be of real value, so I\'d be grateful for a moderator\'s definitive answer on that one.

ID: 23195 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7233
Credit: 23,154,247
RAC: 170
Message 23208 - Posted: 18 Jun 2006, 19:50:10 UTC

\'Sulphur\' models were a short, 4-6 months, project, starting in about October last year, to get values of sulphates for starting the Coupled Ocean models. Which is why the early sulphur models had a very short deadline.
The coupled Ocean models started on February 14th, as TCMs, Transient Coupled Models.

Whether or not people wish to continue processing sulphur models is up to them.
As you are running very slowly on a laptop, it may not be worth it. Phase one was set up to return extra data, which is/was valuable, and your first model has successfully passed this point.

Moderators are no more knowledgable in this matter than other crunchers, but Carl said in the \'spinup\' project that the people running spinup, (a special project to get more values to start TCM), should move to TCM when they finished.

ID: 23208 · Report as offensive     Reply Quote
old_user162591

Send message
Joined: 15 Feb 06
Posts: 3
Credit: 5,568
RAC: 0
Message 23212 - Posted: 18 Jun 2006, 21:08:44 UTC - in response to Message 23208.  

\'Sulphur\' models were a short, 4-6 months, project, starting in about October last year, to get values of sulphates for starting the Coupled Ocean models. Which is why the early sulphur models had a very short deadline.
The coupled Ocean models started on February 14th, as TCMs, Transient Coupled Models.

Whether or not people wish to continue processing sulphur models is up to them.
As you are running very slowly on a laptop, it may not be worth it. Phase one was set up to return extra data, which is/was valuable, and your first model has successfully passed this point.

Moderators are no more knowledgable in this matter than other crunchers, but Carl said in the \'spinup\' project that the people running spinup, (a special project to get more values to start TCM), should move to TCM when they finished.


Thanks, Les. It looks as though I should cut my losses and start with another project.
ID: 23212 · Report as offensive     Reply Quote
Profile Pooh Bear 27
Avatar

Send message
Joined: 5 Feb 05
Posts: 465
Credit: 1,914,189
RAC: 0
Message 23213 - Posted: 18 Jun 2006, 21:28:37 UTC - in response to Message 23212.  

Thanks, Les. It looks as though I should cut my losses and start with another project.

No losses, really. You still got credit for the portion you did do. I killed one Sulphur unit because of this, and never looked back. It was about 50% done, and when I read they only needed the first part of it, I was fine with that knowledge.

Now I am busy with 4 coupled models. They are going to take a while, but that\'s the reason we are here, to predict the weather, and hopefully figure out what is coming.


ID: 23213 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7233
Credit: 23,154,247
RAC: 170
Message 23214 - Posted: 18 Jun 2006, 22:23:17 UTC

I was about 10 trickles, 1 week, short of finishing my one and only sulphur when I started Spinup last November. When I finished that on April 1, I went back and finished the sulphur model.

For others reading this, it\'s all about where people are up to, and how they feel about the time and electricity needed to finish.

Everyone needs to be aware that the new TCM\'s are a 160 year model, taking at least 3 months running 24/7 on very fast computers, and perhaps over a year for \'part-timers\', and those with slow computers.
So BACK UP your BOINC folder at least once a week! Otherwise you\'re asking for trouble.

ID: 23214 · Report as offensive     Reply Quote

Questions and Answers : Preferences : Time to Completion and Sulphur Model

©2020 climateprediction.net