Register | Sign In


Understanding through Discussion


EvC Forum active members: 65 (9162 total)
2 online now:
Newest Member: popoi
Post Volume: Total: 915,817 Year: 3,074/9,624 Month: 919/1,588 Week: 102/223 Day: 0/13 Hour: 0/0


Thread  Details

Email This Thread
Newer Topic | Older Topic
  
Author Topic:   TeamEvC: Climate Prediction.
ohnhai
Member (Idle past 5162 days)
Posts: 649
From: Melbourne, Australia
Joined: 11-17-2004


Message 1 of 20 (347919)
09-10-2006 9:55 AM


I noticed a link in Kuseru's Sig
ClimatePrediction
Basically it's a Distributed computing experiment set up by the BBC to explore the robustness of the Model used by the UK Met office for long range climate predictions.
Anyway if you have a reasonably quick machine and sit for many hours with it idling away while you read EvC or watching DVDs then you might want to consider donating a few free cycles to this project.
Curently there are three of us in the team (Was started by Chris)
Note read the FAQs especially in regard to machine sepcs, how long it will take, how long you normally have your machine on for and quitting early . .. Before connecting to the project.

TeamEvC. join up at Climate Modelling to help improve climate predictions.

Replies to this message:
 Message 2 by kuresu, posted 09-10-2006 3:36 PM ohnhai has replied

  
kuresu
Member (Idle past 2513 days)
Posts: 2544
From: boulder, colorado
Joined: 03-24-2006


Message 2 of 20 (347955)
09-10-2006 3:36 PM
Reply to: Message 1 by ohnhai
09-10-2006 9:55 AM


thanks for joining the team
beat me to creating this thread, but who cares.
oh, by the way. . .
kuresu = chris
one warning to all--these projects take a while, unlike the TeamEvC on the world community grid--going up against HIV/AIDS.
if I understand it correctly, everyone on the team works on the same project--in this case climate prediction. the one currently running on my computer is running through the weather of the 20th century. After running all night, it made it from february 1921 to october 1921. Also, if anyone has any ideas as to how to make this more effiecient, let me know. Being the founder, anyone who joins the team (and please do) uses my preferences for the project. So . . .
anywho
please join up people
(especially you lurkers--all 130 of you)
abe:
another cool feature of this project--you can actually see the model running, and switch between temp, pressure, cloud and rain. (at least, you all should be able to, depending on how you get it setup)
you can't do that with the HIV/AIDS project.
Edited by kuresu, : No reason given.

Want to help give back to the world community? Did you know that your computer can help? Join the newest TeamEvC URL=http://climateapps2.oucs.ox.ac.uk/cpdnboinc/team_display.php Modelling[/URL] to help improve climate predictions for a better tomorrow.

This message is a reply to:
 Message 1 by ohnhai, posted 09-10-2006 9:55 AM ohnhai has replied

Replies to this message:
 Message 3 by Sonne, posted 09-10-2006 5:41 PM kuresu has replied
 Message 7 by ohnhai, posted 09-10-2006 11:00 PM kuresu has not replied

  
Sonne
Member (Idle past 5930 days)
Posts: 58
Joined: 05-20-2006


Message 3 of 20 (347976)
09-10-2006 5:41 PM
Reply to: Message 2 by kuresu
09-10-2006 3:36 PM


It looks very interesting. I'm having trouble getting the model to work, however. This is probably due to my OS though - will keep investigating.
After running all night, it made it from february 1921 to october 1921.
Where does it show the timeline?

This message is a reply to:
 Message 2 by kuresu, posted 09-10-2006 3:36 PM kuresu has replied

Replies to this message:
 Message 4 by kuresu, posted 09-10-2006 8:36 PM Sonne has replied

  
kuresu
Member (Idle past 2513 days)
Posts: 2544
From: boulder, colorado
Joined: 03-24-2006


Message 4 of 20 (347994)
09-10-2006 8:36 PM
Reply to: Message 3 by Sonne
09-10-2006 5:41 PM


the model can be shown if you pick the shared or single install for when the program runs.
to access the model, click on the "tasks" tab. then hit "show graphic". timeline is up at the top of that screen. hope that helps
mine just hit 1922.
one correction--the project I'm working on goes from 1920-2080.
just checked the member's list--we've a total of five people.
Edited by kuresu, : No reason given.

Want to help give back to the world community? Did you know that your computer can help? Join the newest TeamEvC Climate Modelling to help improve climate predictions for a better tomorrow.

This message is a reply to:
 Message 3 by Sonne, posted 09-10-2006 5:41 PM Sonne has replied

Replies to this message:
 Message 5 by Sonne, posted 09-10-2006 9:40 PM kuresu has not replied

  
Sonne
Member (Idle past 5930 days)
Posts: 58
Joined: 05-20-2006


Message 5 of 20 (348005)
09-10-2006 9:40 PM
Reply to: Message 4 by kuresu
09-10-2006 8:36 PM


Still can't view the graphics, maybe an xorg problem... If any other linux users get it working can you please let me know how you did it?
Or maybe I'm missing something obvious...

This message is a reply to:
 Message 4 by kuresu, posted 09-10-2006 8:36 PM kuresu has not replied

Replies to this message:
 Message 6 by ohnhai, posted 09-10-2006 10:54 PM Sonne has not replied

  
ohnhai
Member (Idle past 5162 days)
Posts: 649
From: Melbourne, Australia
Joined: 11-17-2004


Message 6 of 20 (348012)
09-10-2006 10:54 PM
Reply to: Message 5 by Sonne
09-10-2006 9:40 PM


note sure. are you running it as an alpication or a service? I think I reas somewhere the graphics are not available if you run it as a service (could be just a windows thing)

This message is a reply to:
 Message 5 by Sonne, posted 09-10-2006 9:40 PM Sonne has not replied

  
ohnhai
Member (Idle past 5162 days)
Posts: 649
From: Melbourne, Australia
Joined: 11-17-2004


Message 7 of 20 (348013)
09-10-2006 11:00 PM
Reply to: Message 2 by kuresu
09-10-2006 3:36 PM


one warning to all--these projects take a while,
Yup. for a quick machine the runtime could well be 3-4 months this is because for each processor you tell it to use they assign a whole model run (splitting a highly complex model like this across mutliple machines would cause way too much net trafic.)
As my machine is dual core I am currently running two models. getting 2.20 s/TS for the first and around 4 s/TS for the second.

This message is a reply to:
 Message 2 by kuresu, posted 09-10-2006 3:36 PM kuresu has not replied

  
ohnhai
Member (Idle past 5162 days)
Posts: 649
From: Melbourne, Australia
Joined: 11-17-2004


Message 8 of 20 (349190)
09-14-2006 11:34 PM


Then There were three ...
Dang Lost one..... We are down to three active members. we could do with your help... (currently #469 in the team rankings...)
Come join our team.

Replies to this message:
 Message 9 by jar, posted 09-14-2006 11:43 PM ohnhai has not replied
 Message 16 by Sonne, posted 09-15-2006 7:59 PM ohnhai has not replied

  
jar
Member (Idle past 394 days)
Posts: 34026
From: Texas!!
Joined: 04-20-2004


Message 9 of 20 (349193)
09-14-2006 11:43 PM
Reply to: Message 8 by ohnhai
09-14-2006 11:34 PM


Re: Then There were three ...
Had to drop out. Sorry. The sucker took over the BOINC app and had two of them suckers running with two more waiting. It was taking all the time from the Grid project.

Aslan is not a Tame Lion

This message is a reply to:
 Message 8 by ohnhai, posted 09-14-2006 11:34 PM ohnhai has not replied

Replies to this message:
 Message 10 by mick, posted 09-15-2006 1:06 AM jar has not replied

  
mick
Member (Idle past 4986 days)
Posts: 913
Joined: 02-17-2005


Message 10 of 20 (349204)
09-15-2006 1:06 AM
Reply to: Message 9 by jar
09-14-2006 11:43 PM


Re: Then There were three ...
Yes, unless it's possible to run alongside the grid, i will also have to stick with the HIV project.
Mick

This message is a reply to:
 Message 9 by jar, posted 09-14-2006 11:43 PM jar has not replied

  
ohnhai
Member (Idle past 5162 days)
Posts: 649
From: Melbourne, Australia
Joined: 11-17-2004


Message 11 of 20 (349205)
09-15-2006 1:08 AM


Light weights all.. Buy another machine

Replies to this message:
 Message 12 by mick, posted 09-15-2006 1:15 AM ohnhai has not replied

  
mick
Member (Idle past 4986 days)
Posts: 913
Joined: 02-17-2005


Message 12 of 20 (349207)
09-15-2006 1:15 AM
Reply to: Message 11 by ohnhai
09-15-2006 1:08 AM


ok okay I guess I will put it on my work computer tomorrow and keep the HIV thing running on my laptop.

This message is a reply to:
 Message 11 by ohnhai, posted 09-15-2006 1:08 AM ohnhai has not replied

Replies to this message:
 Message 13 by kuresu, posted 09-15-2006 4:08 PM mick has not replied

  
kuresu
Member (Idle past 2513 days)
Posts: 2544
From: boulder, colorado
Joined: 03-24-2006


Message 13 of 20 (349364)
09-15-2006 4:08 PM
Reply to: Message 12 by mick
09-15-2006 1:15 AM


I've actually got the grid (HIV) and the boinc(climate) running at the same time. granted, I've set it so that the grid only consumes 60% of my processing power. I haven't had any problems yet.
Oh, Jar.
This morning I found that I had two climate models, one running. THe only reason I can think this happened is that I had to suspend the model (my game was do some serious freezing up last night--damn graphics card). When I turned on my computer, it had automatically loaded a new climate model and was starting to run it. I aborted that one and got back to my original one. Maybe something similar happened to yours?
thanks for joining, mick

Want to help give back to the world community? Did you know that your computer can help? Join the newest TeamEvC Climate Modelling to help improve climate predictions for a better tomorrow.

This message is a reply to:
 Message 12 by mick, posted 09-15-2006 1:15 AM mick has not replied

Replies to this message:
 Message 14 by kuresu, posted 09-15-2006 7:19 PM kuresu has not replied

  
kuresu
Member (Idle past 2513 days)
Posts: 2544
From: boulder, colorado
Joined: 03-24-2006


Message 14 of 20 (349452)
09-15-2006 7:19 PM
Reply to: Message 13 by kuresu
09-15-2006 4:08 PM


okay--that program just loves to screw with you. If you suspend any model that's running, it will pretty much find a new one when you turn your computer on. When I aborted the new one, it started on a next one. It keeps on unless you have one running. Once I put the original model back on run, it stopped it's download spree. (so long as you have an instantaneous internet connection, like I've got where I'm currently at).
anyone have a clue why it automatically does this, and is there a way to turn that off?

Want to help give back to the world community? Did you know that your computer can help? Join the newest TeamEvC Climate Modelling to help improve climate predictions for a better tomorrow.

This message is a reply to:
 Message 13 by kuresu, posted 09-15-2006 4:08 PM kuresu has not replied

Replies to this message:
 Message 15 by nwr, posted 09-15-2006 7:53 PM kuresu has not replied
 Message 17 by ohnhai, posted 09-15-2006 8:55 PM kuresu has not replied

  
nwr
Member
Posts: 6408
From: Geneva, Illinois
Joined: 08-08-2005
Member Rating: 5.1


Message 15 of 20 (349460)
09-15-2006 7:53 PM
Reply to: Message 14 by kuresu
09-15-2006 7:19 PM


I would guess that it has to be that way.
If you stop the computation on one model, then the distributed processing is going to move that computation to somebody else's computer. And when you restart, it is going to feed you the next thing that needs to be computed.

This message is a reply to:
 Message 14 by kuresu, posted 09-15-2006 7:19 PM kuresu has not replied

  
Newer Topic | Older Topic
Jump to:


Copyright 2001-2023 by EvC Forum, All Rights Reserved

™ Version 4.2
Innovative software from Qwixotic © 2024