Message boards : Number crunching : New region?
Message board moderation
Previous · 1 · 2 · 3 · 4 · Next
Author | Message |
---|---|
Send message Joined: 22 Feb 06 Posts: 491 Credit: 31,037,259 RAC: 14,576 |
Estimated time on my machine is about 13days (3.5GHz i5). |
Send message Joined: 16 Jan 10 Posts: 1084 Credit: 7,827,799 RAC: 5,038 |
Estimated time on my machine is about 13days (3.5GHz i5). Based on the first CAFR trickle from my model, the GFLOPS is overestimated by a factor of 2.8, so divide your 13 days by 2.8 ... |
Send message Joined: 22 Feb 06 Posts: 491 Credit: 31,037,259 RAC: 14,576 |
10% completed after 18hours - so about another 6days to go? |
Send message Joined: 15 May 09 Posts: 4540 Credit: 19,039,635 RAC: 18,944 |
21 days total estimated here. (by my calculations rather than those of BOINC which is telling me my cafr task has 121 days still to go!) |
Send message Joined: 16 Jan 10 Posts: 1084 Credit: 7,827,799 RAC: 5,038 |
[astroWX wrote:]Eventually, all restart files will lose their sequence numbers and be identified with ".restart". This scheme has now been adopted for the new cafr model, thus: wah2_cafr25_a06s_201312_13_468_010778952_1_r452985916_out.zip wah2_cafr25_a06s_201312_13_468_010778952_1_r452985916_restart.zip The out Zip file contains the error and other text reports that used to be bundled with each Zip. I wondered why a properly spun-up climate model would produce Zip files that increased in size, but the reason was those additional text files. The cafr monthly uploads are now reassuringly the same size throughout the model run. |
Send message Joined: 16 Jan 10 Posts: 1084 Credit: 7,827,799 RAC: 5,038 |
Though based only on one completion of mine, the new WAH2 CAFR25 model does indeed have a very pessimistic GFLOPS estimate: Region..Resolution..Months..GFLOPS........Normalised GFLOPS/sec AFR.......50...............13..........1,924,645......1.58 CAFR......25...............13.........5,222,246......2.78 CAM.......50...............13............849,559......0.94 DUB.......25...............13..........3,327,828......0.93 EU(r)......50...............13............898,721......1.12 EU.........25...............12..........1,903,330......0.86 EU.........25...............13..........2,061,502......0.88 MEX.......50...............12............784,376.......0.91 MEX.......50...............13............849,559.......0.92 MEX.......25...............13..........4,172,902......n/a PNW......25...............16..........1,277,597.......n/a PNW......25...............24..........1,915,068.......1.19 PNW......25...............49..........3,907,164.......n/a SAM......50...............13..........2,147,398.......1.68 SAS.......50................8.............921,921.......1.12 SAS.......50...............13..........1,495,731.......1.12 WUS......25.................9..........n/a.................n/a WUS......25...............24..........n/a.................n/a In a mix of models, a model with a "normalised GFLOPS/sec" greater than 1.00 will finish earlier than expected. The new-ish cam50 has an appropriate GFLOPS estimate (it is really a Mexico model). The new eu50r is likewise alright. |
Send message Joined: 22 Feb 06 Posts: 491 Credit: 31,037,259 RAC: 14,576 |
Updated progress about 58% completed after 4 1/2 days on my 3 CAFR25 models. Original BOINC estimate was 13days. |
Send message Joined: 16 Jan 10 Posts: 1084 Credit: 7,827,799 RAC: 5,038 |
The new work reported elsewhere includes some Western United States at 25 km resolution (WUS25). These have appeared before in 9 and 24 month versions but I managed to get one of the new 25 month versions this time, so here is a comparison with the Pacific North-West area. |
Send message Joined: 16 Jan 10 Posts: 1084 Credit: 7,827,799 RAC: 5,038 |
A new type of WAH2 has been and gone today, such as wah2_global_t000_201112_1_480_010799487. |
Send message Joined: 16 Jan 10 Posts: 1084 Credit: 7,827,799 RAC: 5,038 |
A new region has appeared, briefly, over the last few days - East Asia (EAS50). Thanks to Dave Jackson for the Zip file from which the area image is derived. |
Send message Joined: 15 May 09 Posts: 4540 Credit: 19,039,635 RAC: 18,944 |
And another NAWA about 1/10 task for each member with recent credit at a guess! |
Send message Joined: 9 Sep 04 Posts: 228 Credit: 30,756,611 RAC: 3,303 |
And I got three tasks. But still no change to see at server status. UUuuuuuuuuuuuuuuups, 1040 tasks shown there.... Have a nice day, bonsai911 |
Send message Joined: 15 May 09 Posts: 4540 Credit: 19,039,635 RAC: 18,944 |
But still no change to see at server status. It seems to only get updated about every 2-4 hours at the moment. I haven't been watching it closely enough to be sure if the interval is regular or not. |
Send message Joined: 16 Jan 10 Posts: 1084 Credit: 7,827,799 RAC: 5,038 |
The new batch mentioned in the preceding posts is another Africa/Asia variant - NAWA25 (North Africa/West Asia?). Thanks to Dave Jackson for the Zip file from which the area image is derived, and the GFLOPS estimate. |
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
Regions NAWA25 is the last one in the descriptions. |
Send message Joined: 16 Jan 10 Posts: 1084 Credit: 7,827,799 RAC: 5,038 |
My second WAH2 CAFR25 model failed so the very pessimistic GFLOPS estimate stands for the moment; my first pair of WUS25/25 completions is also among the high values: Region..Resolution..Months..GFLOPS........Normalised GFLOPS/sec AFR.......50...............13..........1,924,645......1.59 CAFR.....25...............13..........5,222,246......2.76 CAM......50...............13............849,559.......0.93 DUB......25...............13..........3,327,828......0.93 EAS.......50...............12..........1,777,602......n/a EU.........25...............12..........1,903,330......0.86 EU.........25...............13..........2,061,502......0.88 EU(r)......50...............13............898,721......1.12 MEX.......50...............12............784,376.......0.91 MEX.......50...............13............849,559.......0.92 MEX.......25...............13..........4,172,902......n/a NAWA....25...............12..........n/a................n/a NAWA....25...............13..........3,327,828......n/a PNW......25...............16..........1,277,597.......n/a PNW......25...............24..........1,915,068.......1.20 PNW......25...............48..........n/a................n/a PNW......25...............49..........3,907,164.......n/a SAM......50...............13..........2,147,398.......1.67 SAS.......50................8.............921,921.......1.12 SAS.......50...............13..........1,495,731.......1.12 WUS......25.................9..........n/a.................n/a WUS......25...............24..........n/a.................n/a WUS......25...............25..........3,291,506.......2.05 In a mix of models, a model with a "normalised GFLOPS/sec" greater than 1.00 will finish earlier than expected; the high values are excluded from calculation of the average. The recent NAWA25/12 (batch #494) is available on Mac, which is a welcome return, though I didn't get any. My Linux machine has failed so the mix of Windows/Mac/Linux machines used for these estimates is no longer available to reduce any platform bias. |
Send message Joined: 22 Feb 06 Posts: 491 Credit: 31,037,259 RAC: 14,576 |
I got some new eu50r tasks today which appear to be 3 month runs with a estiamte of 209,165Gflops. |
Send message Joined: 16 Jan 10 Posts: 1084 Credit: 7,827,799 RAC: 5,038 |
I got some new eu50r tasks today which appear to be 3 month runs with a estiamte of 209,165Gflops. ... which is interesting, because the GFLOPS value is now the same for WAH2 and HADAM3P(EU), so a comparison is then possible. |
Send message Joined: 16 Jan 10 Posts: 1084 Credit: 7,827,799 RAC: 5,038 |
I got some new eu50r tasks today which appear to be 3 month runs with a estiamte of 209,165Gflops. The WAH2 variant, at least for batch #496, runs decidedly quicker than expected (normalised GFLOPS/sec = 1.51), both relative to the collection of other WAH2 models and comparable HADAM3P models (e.g. batch #469). |
Send message Joined: 22 Feb 06 Posts: 491 Credit: 31,037,259 RAC: 14,576 |
The latest batch of new ones are eas50 with an apparent run of 13 months and an estimated size of 1,925,325Gflops. |
©2024 cpdn.org