climateprediction.net home page
New region?
New region?
log in

Advanced search

Message boards : Number crunching : New region?

Previous · 1 · 2 · 3 · 4 · Next
Author Message
Profile Iain Inglis
Volunteer moderator
Send message
Joined: 16 Jan 10
Posts: 959
Credit: 1,964,466
RAC: 9,794
Message 54884 - Posted: 5 Oct 2016, 16:45:26 UTC

Thanks, Byron.

The "49" should signify 49 monthly Zips, plus one restart Zip. The restart Zip will be Zip #50; however, if the pattern for other odd-numbered models is followed then it will be produced at the same time as Zip #48.

I assume Zip #49 is produced in order to allow some quantity to be interpolated in the month before; or maybe someone's checking out whether Month #1 of the next model in the sequence is the same as the Month #49 in the current model.

Or maybe they're weekly Zips!

Profile astroWX
Volunteer moderator
Send message
Joined: 5 Aug 04
Posts: 1485
Credit: 89,881,625
RAC: 24,567
Message 54887 - Posted: 6 Oct 2016, 18:06:26 UTC

They are, in fact, monthly.

Eventually, all restart files will lose their sequence numbers and be identified with ".restart".
____________
"We have met the enemy and he is us." -- Pogo
Greetings from coastal Washington state, the scenic US Pacific Northwest.

Profile JIM
Send message
Joined: 31 Dec 07
Posts: 1071
Credit: 18,636,429
RAC: 5,368
Message 54920 - Posted: 12 Oct 2016, 15:21:23 UTC

Is there a new project? I just received four tasks marked wah2_cam50. I assume that the 50 is 50 months, but, I don’t remember ever seeing a WU designated “_cam.”
____________

Les Bayliss
Volunteer moderator
Send message
Joined: 5 Sep 04
Posts: 6781
Credit: 20,534,340
RAC: 8,056
Message 54921 - Posted: 12 Oct 2016, 15:44:29 UTC - in response to Message 54920.

Hi Jim

50 is the resolution of the "regional model", in this case 50Km.
cam is Central America.

The run length is 13 months.

Profile JIM
Send message
Joined: 31 Dec 07
Posts: 1071
Credit: 18,636,429
RAC: 5,368
Message 54922 - Posted: 12 Oct 2016, 17:25:54 UTC - in response to Message 54921.

Thanks Les. I should have realized the 50 was resolution and not duration.
____________

Profile Iain Inglis
Volunteer moderator
Send message
Joined: 16 Jan 10
Posts: 959
Credit: 1,964,466
RAC: 9,794
Message 54925 - Posted: 12 Oct 2016, 20:34:26 UTC

Thanks for the heads-up, Jim. Managed as a result to get a couple - not bad from a hotel room in Rome!

Profile Alan K
Send message
Joined: 22 Feb 06
Posts: 263
Credit: 13,330,471
RAC: 10,279
Message 54926 - Posted: 12 Oct 2016, 22:21:56 UTC - in response to Message 54925.

Ooops! Managed to pick up 8!!

Profile Iain Inglis
Volunteer moderator
Send message
Joined: 16 Jan 10
Posts: 959
Credit: 1,964,466
RAC: 9,794
Message 54945 - Posted: 17 Oct 2016, 12:34:46 UTC
Last modified: 22 Mar 2017, 10:25:40 UTC

Here is the area covered by the recently released Central America 50 km WAH2 model (CAM50):



[Edit: Which is the same as MEX50.]

Profile Iain Inglis
Volunteer moderator
Send message
Joined: 16 Jan 10
Posts: 959
Credit: 1,964,466
RAC: 9,794
Message 54950 - Posted: 17 Oct 2016, 18:15:31 UTC

... and some WAH2 EU50r models, which are new to me - e.g. wah2_eu50r_b000_198512_13_460_010762307.

Profile Byron Leigh Hatch @ team Carl Sagan
Avatar
Send message
Joined: 17 Aug 04
Posts: 279
Credit: 43,022,075
RAC: 0
Message 54952 - Posted: 18 Oct 2016, 4:38:55 UTC - in response to Message 54950.

Hello Iain and hello everyone,
yes my computer is now crunching,
14 of these of these wah2_eu50r_ models now,
according to BOINC ... the 14 ... wah2_eu50r_ models should,
finish in two (2) days and 9 hours, or approximatly by 20 October 2016.

an example below:

wah2_eu50r_b175_201312_13_460_010763860_0

Profile Iain Inglis
Volunteer moderator
Send message
Joined: 16 Jan 10
Posts: 959
Credit: 1,964,466
RAC: 9,794
Message 54953 - Posted: 18 Oct 2016, 9:59:30 UTC

Thanks, Byron. That GFLOPS estimate has now been added to the list. I'll just have to wait to get a model to run ...

Profile Iain Inglis
Volunteer moderator
Send message
Joined: 16 Jan 10
Posts: 959
Credit: 1,964,466
RAC: 9,794
Message 54958 - Posted: 18 Oct 2016, 16:54:23 UTC - in response to Message 54953.
Last modified: 18 Oct 2016, 16:58:52 UTC

There's a flurry of batch explanations on the front page, including EU50r - http://www.climateprediction.net.

[Edit: I can't create an image of the rotated EU region without a model - and it looks like I missed out this time, but the batch is a "spin-up" so there will no doubt be more at some point.]

Profile Iain Inglis
Volunteer moderator
Send message
Joined: 16 Jan 10
Posts: 959
Credit: 1,964,466
RAC: 9,794
Message 54962 - Posted: 18 Oct 2016, 19:26:15 UTC
Last modified: 22 Mar 2017, 10:25:54 UTC

Here is the new WAH2 EU50r area, which corresponds rather better with the rotated area in the regions map than the HADAM3P-derived EU50 area earlier in this thread.

John Eric Hopkinson
Send message
Joined: 27 Jan 05
Posts: 74
Credit: 1,028,011
RAC: 0
Message 54975 - Posted: 21 Oct 2016, 1:14:35 UTC - in response to Message 54962.

Iain:
These area maps add a reality dimension to the crunching here and thus add an extra dimension to our interest in these studies. Just how CPDN is flinging these wu's out to the participants is another question, because the ones I receive are literally all over the map. So the distribution concentrates on computer availability rather than who is where. Right?
____________

Profile JIM
Send message
Joined: 31 Dec 07
Posts: 1071
Credit: 18,636,429
RAC: 5,368
Message 54976 - Posted: 21 Oct 2016, 2:15:58 UTC - in response to Message 54975.

It seems that you just get the next task in hopper, regardless of where you live.
____________

Les Bayliss
Volunteer moderator
Send message
Joined: 5 Sep 04
Posts: 6781
Credit: 20,534,340
RAC: 8,056
Message 54977 - Posted: 21 Oct 2016, 3:00:27 UTC

It depends on which research group, looking at which global area, is the next to "drop a batch in the hopper". Along with re-issues, which the server pops back in the hopper as they come up.

So Jim is correct; it's whatever happens to come out of the hopper next.

e.g. with the recent sas50, batch 461, 2 of my computers got only them, but the 3rd computer got 3 of the 461s, and a re-issue of a afr50 batch 459, which is disappointing, as I was hoping for a "full set". And the afr50 is 13 months, compared to the 8 months sas50.

Profile Iain Inglis
Volunteer moderator
Send message
Joined: 16 Jan 10
Posts: 959
Credit: 1,964,466
RAC: 9,794
Message 55125 - Posted: 13 Nov 2016, 19:43:48 UTC

There's a new region designation at 25 km resolution, nawa, described as "over Dubai region" (one month). There are only 10 work units, starting at wah2_nawa25_brar_200712_1_464_010778541.

There are then some Western US, starting at wah2_wus25_t04r_19869_24_466_010778551, this time at 24 months. The application version has been bumped to 8.24.

PS As ever, if anyone has a GFLOPS estimate or, even better, the ZIP download, do send it over and I'll add them to the list and plot the region.

Profile Iain Inglis
Volunteer moderator
Send message
Joined: 16 Jan 10
Posts: 959
Credit: 1,964,466
RAC: 9,794
Message 55146 - Posted: 16 Nov 2016, 14:49:19 UTC
Last modified: 22 Mar 2017, 10:26:06 UTC

The new work reported elsewhere includes a new region, Central Africa, at 25 km resolution.



PS Not quick enough this time to snaffle a mex25.

Profile Alan K
Send message
Joined: 22 Feb 06
Posts: 263
Credit: 13,330,471
RAC: 10,279
Message 55148 - Posted: 16 Nov 2016, 15:08:12 UTC - in response to Message 55125.

Hi Ian. Got a Gflops estimate of 5,222,246 for a CAFR25 model if it helps. Managed to get 3 of them on one machine.

Profile Iain Inglis
Volunteer moderator
Send message
Joined: 16 Jan 10
Posts: 959
Credit: 1,964,466
RAC: 9,794
Message 55149 - Posted: 16 Nov 2016, 15:20:32 UTC - in response to Message 55148.

Hi Ian. Got a Gflops estimate of 5,222,246 for a CAFR25 model if it helps. Managed to get 3 of them on one machine.

Thanks, Alan. It might take a few days by the look of it to see whether it's accurate or not - I've got a couple. My estimate for the recent WAH2 versions for Africa and South America is that they were 60% out (i.e. the GFLOPS should be divided by 1.6). However, the recent Dubai area was pretty much spot on, so the process appears to be a bit variable - or perhaps my machines are variable (though Windows/Mac/Linux comparisons are remarkably consistent).

Previous · 1 · 2 · 3 · 4 · Next

Message boards : Number crunching : New region?


Main page · Your account · Message boards


Copyright © 2019 climateprediction.net