Cross Project Credit Equalization and Adjustment
Message boards :
Number crunching :
Cross Project Credit Equalization and Adjustment
Message board moderation
Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 . . . 8 · Next
| Author | Message |
|---|---|
 Jord  Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 
|
Results for PrimeGrid 5.13, on a P4 2.8GHz, 1GB DDR333, Windows 2000 SP4.
Result in CPU time CC GC
(min:sec)
7 Jun 2006 6:41:59 UTC Over Success Done 0:24:11 3.41 3.46
7 Jun 2006 6:16:00 UTC Over Success Done 0:23:51 3.36 4.59
7 Jun 2006 5:50:21 UTC Over Success Done 0:24:14 3.41 3.37
7 Jun 2006 5:24:50 UTC Over Success Done 0:23:47 3.35 3.01
7 Jun 2006 1:59:17 UTC Over Success Done 0:19:12 2.70 2.50
6 Jun 2006 20:09:33 UTC Over Success Done 0:23:33 3.32 5.00
6 Jun 2006 20:09:33 UTC Over Success Done 0:13:32 1.91 4.83
6 Jun 2006 20:09:33 UTC Over Success Done 0:23:32 3.31 5.30
6 Jun 2006 20:09:33 UTC Over Success Done 0:23:18 3.28 5.41
6 Jun 2006 20:09:33 UTC Over Success Done 0:23:07 3.26 5.28
6 Jun 2006 20:09:33 UTC Over Success Done 0:23:03 3.24 5.32
6 Jun 2006 20:09:33 UTC Over Success Done 0:23:45 3.34 5.14
6 Jun 2006 20:09:33 UTC Over Success Done 0:23:03 3.25 5.38
6 Jun 2006 20:09:33 UTC Over Success Done 0:23:13 3.27 5.39
6 Jun 2006 20:09:33 UTC Over Success Done 0:22:46 3.21 5.35
6 Jun 2006 1:53:15 UTC Over Success Done 0:24:52 3.50 4.89
6 Jun 2006 1:53:14 UTC Over Success Done 0:23:41 3.33 5.00
6 Jun 2006 1:53:14 UTC Over Success Done 0:24:46 3.49 5.60
6 Jun 2006 1:53:14 UTC Over Success Done 0:24:31 3.45 3.70
6 Jun 2006 1:53:14 UTC Over Success Done 0:24:44 3.48 3.78
6 Jun 2006 1:53:13 UTC Over Success Done 0:24:20 3.43 3.76
5 Jun 2006 12:33:58 UTC Over Success Done 0:23:44 3.53 3.70
5 Jun 2006 12:33:57 UTC Over Success Done 0:19:10 2.85 3.41
5 Jun 2006 12:33:57 UTC Over Success Done 0:23:37 3.51 3.44
5 Jun 2006 12:33:57 UTC Over Success Done 0:23:29 3.49 2.96
5 Jun 2006 12:33:57 UTC Over Success Done 0:24:05 3.58 3.03
5 Jun 2006 12:33:57 UTC Over Success Done 0:23:52 3.55 5.14
5 Jun 2006 12:33:57 UTC Over Success Done 0:25:28 3.78 2.66
4 Jun 2006 23:16:02 UTC Over Success Done 0:22:54 3.40 2.88
4 Jun 2006 23:16:02 UTC Over Success Done 0:23:20 3.47 3.84
4 Jun 2006 23:16:02 UTC Over Success Done 0:23:10 3.44 3.43
4 Jun 2006 23:16:02 UTC Over Success Done 0:23:25 3.48 2.94
4 Jun 2006 23:16:02 UTC Over Success Done 0:24:14 3.60 3.95
4 Jun 2006 23:16:01 UTC Over Success Done 0:23:46 3.53 2.99
3 Jun 2006 11:44:22 UTC Over Success Done 0:23:48 3.54 3.12
3 Jun 2006 7:37:14 UTC Over Success Done 0:23:30 3.49 2.58
3 Jun 2006 4:50:27 UTC Over Success Done 0:24:17 3.61 2.65
2 Jun 2006 12:26:56 UTC Over Success Done 0:23:01 3.42 4.13
2 Jun 2006 12:03:32 UTC Over Success Done 0:22:31 3.35 2.30
2 Jun 2006 10:24:14 UTC Over Success Done 0:23:07 3.44 2.97
3 Jun 2006 4:50:27 UTC Over Success Done 0:25:10 3.74 3.05
3 Jun 2006 0:44:46 UTC Over Success Done 0:24:34 3.65 3.60
2 Jun 2006 22:15:10 UTC Over Success Done 0:29:11 4.34 3.18
2 Jun 2006 19:59:30 UTC Over Success Done 0:24:37 3.66 2.84
Primegrid has a quorum of 2. Granted credit is the claimed credit of two hosts added up and divided by two. Total time: 6h 05m 45sec Total CC: 146.22 Cobblestone/hour claimed: 0.04 I won't add up the other list, as it's deluted by the adding up and divided by 2. ID: 329587 ·  |
Pappa  Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 
|
Eric here are some of my latest results from Predictor. Peter Can You give me a URL to your preditor Results? Similar to this http://setiathome.berkeley.edu/show_user.php?userid=1105331 (your Seti BOINC) Edit This allows a look at the same "machine" in a Cross Project... Then I can cut an paste in the spreadsheet for Eric. This allows for faster processing of what we have been trying to get accomplished for weeks. It is Sad that some many things are in the way... Now is the time for people to get back to work making Seti Better and healing from the Frustration that has been expressed. I will be happy to trim the monster Spreadsheet that I have and leave an example page of what to put in... If you are intereseted in the spreadsheet please email al.setiboinc (at) gmail.com Pappa Please consider a Donation to the Seti Project. ID: 329742 · |
|
Clyde C. Phillips, III Send message Joined: 2 Aug 00 Posts: 1851 Credit: 5,955,047 RAC: 0
|
When comparing credit per hour across the workunits of Seti (by itself and with those of all other projects isn't claimed credit the thing to use and not granted credit? It seems like using granted credit would add an irrelevant variable. ID: 329866 · |
|
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0
|
Actually it is a number that has to be looked at when there is a "qourum" involved... So to have the data is a good thing... It also helps when people want to understand why "my credit goes up and down" "cyclic" it is the quorum at work... Seti Enhanced you may see some differences do the fact that people are running 4.xx BOINC CORE Clients. I have one machine that is running it for that very purpose (data collection)... When comparing credit per hour across the workunits of Seti (by itself and with those of all other projects isn't claimed credit the thing to use and not granted credit? It seems like using granted credit would add an irrelevant variable. Please consider a Donation to the Seti Project. ID: 329994 · |
Saenger  Send message Joined: 3 Apr 99 Posts: 2452 Credit: 33,281 RAC: 0 
|
Results from my current machine (Seti has just restarted, have to wait yet for the first enhanced to get Credit) I'm running stock client 5.4.9 and stock applications. Projekt WU Time [sec] Claimed Granted C/h claim QMC 113783 46442,91 64,15 64,15 4,97 103948 39763,02 58,67 58,67 5,31 93749 40328,41 57,89 57,89 5,17 Einst. 8426733 5927,13 8,19 12,48 4,97 8995212 6175,75 9,13 9,13 5,32 8994504 6021,35 8,9 8,9 5,32 8994501 6041,21 8,93 8,93 5,32 8928122 19161,4 28,34 44,25 5,32 LHC 1403794 148,87 0,22 0,31 5,32 1394109 100,83 0,15 0,16 5,36 1390575 3604,57 5,33 7,11 5,32 Malaria 184875 3667,31 5,42 5,42 5,32 184867 2977,38 4,4 7,18 5,32 184866 2817,54 4,17 8,84 5,33 184860 3410,29 5,04 9,99 5,32 184859 3410,06 5,04 5,04 5,32 184856 3714,04 5,49 7,88 5,32 184852 3666,22 5,42 8,35 5,32 184851 3585,19 5,3 6,52 5,32 184848 3622,93 5,36 6,52 5,33 184846 3598,57 5,32 6,89 5,32 184842 3592,22 5,31 6,66 5,32 184837 3243,89 4,8 7,19 5,33 181608 3143,06 4,65 4,65 5,33 181605 3053,52 4,52 7,69 5,33 181604 3055,58 4,52 5,93 5,33 181600 2875,97 4,25 4,03 5,32 181423 2963,26 4,38 4,38 5,32 181414 3032,86 4,49 4,49 5,33 173678 3375,61 4,99 4,99 5,32 181592 2707,33 4 5,9 5,32 181591 2624,1 3,88 5,87 5,32 181588 2621,31 3,88 3,88 5,33 181575 3300,5 4,88 8,15 5,32 181572 3154,66 4,67 6,23 5,33 176644 2707,91 4 5,24 5,32 176633 2857,28 4,22 6,29 5,32 170954 2741,58 4,05 7,81 5,32 Pirates 280568 18,82 0,03 0,03 5,74 280567 18,82 0,03 0,03 5,74 280566 18,46 0,03 0,03 5,85 Predic. 7653692 4815,14 7,29 8,67 5,45 7653689 4816,84 7,3 10,58 5,46 7653571 4824,76 7,31 7,98 5,45 7653553 5332,81 8,08 6,06 5,45 7653465 4921,21 7,45 5,85 5,45 7653162 5084,54 7,7 6,18 5,45 7439099 5165,59 8,37 8,37 5,83 7438916 5493,9 8,91 8,91 5,84 7438785 5529,68 8,96 9,43 5,83 7436677 5386,5 8,73 7,32 5,83 7371679 5125,96 8,31 8,31 5,84 7371458 5173,62 8,39 7,5 5,84 7371432 5198,8 8,43 7,95 5,84 Rosetta 18955370 10775,36 14,88 14,88 4,97 18843109 10785,86 14,9 14,9 4,97 18694740 10504,44 15,53 15,53 5,32 18402566 10483 15,5 15,5 5,32 Simap 1334426 4646,59 6,87 7,47 5,32 1332164 4581,23 6,77 9,37 5,32 Sztaki 827768 1624,76 2,24 2,24 4,96 826431 4558,16 6,3 12,27 4,98 825751 1466,97 2,03 2,63 4,98 821507 2316,34 3,43 3,43 5,33 813856 3488,94 5,16 5,41 5,32 Xtreme 1201161 594,58 0,88 0,88 5,33 1201160 595,22 0,88 0,88 5,32 1201159 595,29 0,88 0,88 5,32 1201127 593,8 0,88 0,88 5,34 1201122 563,72 0,83 0,83 5,30 WCG ex103_05 33804 47 47 5,01 ex103_1F 39708 55 55 4,99 Gruesse vom Saenger For questions about Boinc look in the BOINC-Wiki ID: 330051 · |
Astro  Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
Eric, I only had two SE wus to use but on my AMD643700's, but "claimed" credit comes out. Eric, above is what my claimed credit was for my AMD64 3700's, at the time I din't have much SE on the results page to work with, so I started collecting. I now have some, but it seems you're in a hurry, so I'll give you wnat I've gotten. All my reports are stock boinc 5.4.9 and stock apps here's SE for my AMD64 3700 Sandiego ar=0.013220 336390655 80484093 4 Jun 2006 0:40:55 UTC 6 Jun 2006 3:13:41 UTC Over Success Done 11,638.30 26.37 26.45 noisey .0023 cc/sec ar=0.43157 335448619 80426796 2 Jun 2006 2:48:51 UTC 3 Jun 2006 5:43:31 UTC Over Success Done 13,861.06 61.47 61.47 .0044 cc/sec ar=0.438683 337716961 80951273 6 Jun 2006 3:10:10 UTC 6 Jun 2006 21:41:37 UTC Over Success Done 15,031.19 61.27 61.27 .0041 cc/sec ar=0.484803 336390580 80461844 4 Jun 2006 0:40:55 UTC 5 Jun 2006 1:37:40 UTC Over Success Done 5,975.72 25.33 25.45 noisey .0043 cc/sec ar=0.566300 335838816 80517358 2 Jun 2006 21:49:50 UTC 4 Jun 2006 0:40:55 UTC Over Success Done 11,985.75 49.92 50.27 .0042 cc/sec ar=0.649373 336390654 80511761 4 Jun 2006 0:40:55 UTC 4 Jun 2006 8:35:37 UTC Over Success Done 11,086.81 45.87 45.87 .0041 cc/sec ar 0.896279 335448738 80426802 2 Jun 2006 2:48:51 UTC 2 Jun 2006 11:47:41 UTC Over Success Done 8,105.11 29.33 29.33 .0036 cc/sec ID: 330109 · |
|
Astro Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
my Gateway amd64 3700 Mobile processor 754 socket doesn't claim as much. Stock boinc, stock apps: it claims .0037 on Ralph, .0037 on Simap, and .0037 on Rosetta ar 0.40973 326268056 78324880 15 May 2006 11:50:09 UTC 23 May 2006 3:13:02 UTC Over Success Done 15,698.73 63.83 63.83 .0041 cc/sec ar=0.426359 335448850 80426836 2 Jun 2006 2:49:32 UTC 4 Jun 2006 0:40:37 UTC Over Success Done 21,674.06 62.4 62.76 .0029 cc/sec 335448813 80426844 2 Jun 2006 2:49:32 UTC 4 Jun 2006 20:24:30 UTC Over Success Done 21,291.61 62.4 62.4 .0029 cc/sec 336849315 80761192 4 Jun 2006 23:33:41 UTC 7 Jun 2006 0:57:44 UTC Over Success Done 20,413.34 60.9 60.9 .0030 cc/sec ar=0.719631 336416500 80654579 4 Jun 2006 1:57:54 UTC 5 Jun 2006 11:16:31 UTC Over Success Done 12,024.27 33.24 33.24 .0028 cc/sec ID: 330120 · |
|
Astro Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
My P4 1.8 is here:it claims .00154 on Rosetta, .00154 on Simap, ar=0.426359 335468441 80431620 2 Jun 2006 3:48:37 UTC 3 Jun 2006 11:16:56 UTC Over Success Done 10,038.34 15.18 15.19 .0015 cc/sec ar=0.536906 336390713 80515140 4 Jun 2006 0:41:06 UTC 5 Jun 2006 18:13:51 UTC Over Success Done 30,551.86 52.6 52.6 .0017 cc/sec ar=1.17637 335468357 80431615 2 Jun 2006 3:48:37 UTC 2 Jun 2006 15:21:46 UTC Over Success Done 11,596.92 19.19 19.19 .0017 cc/sec and my Celeron 500 win98SE, 256 M ram is here: It claims .0007 on simap, and .0007 on Ralph ar=0.664740 335630127 80471322 2 Jun 2006 12:08:08 UTC 7 Jun 2006 4:15:29 UTC Over Success Done 123,146.00 34.5 34.51 .0003 cc/sec so from these figures I'd say seti multiplier is [edit]about right[/edit], but my few machines and the small number of seti samples don't make up the whole picture. my samples on other projects is much more substantial ID: 330128 · |
|
P . P . L . Send message Joined: 7 Jun 03 Posts: 86 Credit: 161,216 RAC: 0 
|
Pappa I didn't think to add that before. http://predictor.scripps.edu/results.php?userid=55467 Sounds like a lot of work. 
ID: 330131 · |
Odysseus  Send message Joined: 26 Jul 99 Posts: 1808 Credit: 6,701,347 RAC: 6 
|
Here are some more data, from a couple of Mac G4/400s (BOINC Menubar 5.2.13). Again I’m noting both claimed and granted credit because of the large discrepancies between the two on E@h & SDG: “Buzz†Application Claimed/h Ratio Granted/h Ratio S@h Enh. v5.13 2.219 1.000 2.219 1.000 Einstein v4.56 1.755 0.791 3.273 1.475 SZTAKI v1.12 1.748 0.788 1.688 0.761 “Rabelais†S@h Enh. v5.13 2.269 1.000 2.269 1.000 Einstein v4.56 1.807 0.796 3.117 1.374 SZTAKI v1.12 1.807 0.796 3.377 1.488Very few S@h Enhanced units were available for the above, but their ARs were near 0.4 so I'm assuming they were reasonably typical. “Buzz†had a somewhat higher than usual rate of validation failure on SDG across this sample, while “Rabelais†was doing especially well in that regard.  ID: 330294 · |
|
Robert Everly Send message Joined: 19 May 99 Posts: 29 Credit: 128,573 RAC: 0
|
Update on my data: host 12619 LHC average 8.32 cr/h and Rosetta 8.31 cr/h wu ar time claimed granted cr/h 80495014 ar=0.369845 43381.75 69.90 pending 5.80 79007363 ar=0.426485 8617.90 14.28 14.28 5.96 79989416 ar=0.443414 36807.87 60.68 60.67 5.93 78894992 ar=0.447153 36246.67 60.19 60.18 5.98 80921238 ar=0.468429 35628.88 58.07 58.06 5.87 host 12638 LHC average 7.87 cr/h and Rosetta 8.97 cr/h wu ar time claimed granted cr/h 80491093 ar=0.138029 9770.88 20.40 20.40 7.52 80497887 ar=0.426359 24230.09 62.41 62.40 9.27 80963838 ar=0.433781 24547.67 62.19 62.18 9.12 80580793 ar=0.731393 14208.95 32.89 32.89 8.33 host 2390564 LHC average 5.66 cr/h and Rosetta 5.61 cr/h (with HT on) wu ar time claimed granted cr/h 80572882 ar=0.426360 27553.33 62.40 62.40 8.15 80988003 ar=0.433781 28268.64 62.18 62.18 7.92 81032960 ar=0.433832 28292.27 62.18 7.91 80949437 ar=0.439351 27526.97 61.20 8.00 80542315 ar=0.462402 26269.56 58.93 58.93 8.08 80572873 ar=0.465336 26195.50 58.59 58.59 8.05 80579004 ar=0.512014 24462.88 54.33 8.00 80572893 ar=0.691015 17127.28 33.98 7.14 80528465 ar=0.735905 15815.73 32.47 55.18 7.39 80971741 ar=3.927571 3696.30 12.89 12.89 12.55 80488087 ar=6.349359 4390.03 12.54 12.54 10.28 80488095 ar=6.349359 3361.39 12.54 12.54 13.43 80488103 ar=6.349359 4516.66 12.54 12.54 9.99 What I find interesting is looking at the graphs in excel is my Amd 1900 is underclaiming by about 2 cr/h, the Sempron is right in the ballpark, while the P4 is overclaiming about 2 cr/h. Very odd. Anyone else seeing the same thing? ID: 330659 · |
W-K 666  Send message Joined: 18 May 99 Posts: 19844 Credit: 40,757,560 RAC: 67 
 |
Update on my data:...... Seti does seem to like Intels more than AMD's and Seti also works better with larger L2 cache. But any restriction by the motherboard on memory bandwidth will lower Seti performance. Andy ID: 330806 · |
|
Jochen Send message Joined: 17 Nov 01 Posts: 15 Credit: 1,019,307 RAC: 0
|
I do not see any point in these lists? What are they supposed to proove? Incompetence of the Berkeley staff?!? This is comparing apples with bananas. This will not work out. It might be fair in a cross-project view. But in a SETI internal view, you annilihated any project internal means of comparison. I will follow Chrunch3rs example and stop SETI on my computers. In fact I already started disconnecting the clients. If this is what you intended: Good job! ID: 330931 · |
|
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13985 Credit: 208,696,464 RAC: 304
 |
I do not see any point in these lists? If you read the first post you'll see what the point is. I will follow Chrunch3rs example and stop SETI on my computers. In fact I already started disconnecting the clients. Then why add to the noise here? Grant Darwin NT ID: 330932 · |
Eric Korpela  Send message Joined: 3 Apr 99 Posts: 1383 Credit: 54,506,847 RAC: 60
|
To some extent that is to be expected. One of the key parameters in determining the how much work can actually be done is the ratio of CPU speed to memory bandwidth. In general, CPU speeds have increased faster than memory bus speeds. The way this shows up is that machines with fast CPUs tend to overclaim with respect to the average machine using the old (time x benchmark) credit method. Machines with slow processors tend to underclaim for the same reason. It's a fairly general trend even in the old credit method (if multi-result validation is used) fast machines tended to claim more than they were granted and slow machines tended to claim less than they were granted. For example, if I have a 2.5 GHz machine that uses memory that is only 3x as fast as the memory used in a 500 MHz machine, the benchmarks will say that the 2.5 GHz machine is more than 5X as fast, but the granted credit will probably say it's somewhere between 3 and 5x as fast (probably somewhere around 3.75X as fast). It's been on my list to get memory benchmarks back into BOINC for this very reason. 
ID: 331043 · |
|
Jochen Send message Joined: 17 Nov 01 Posts: 15 Credit: 1,019,307 RAC: 0
|
If you read the first post you'll see what the point is. I did. May I propose that you read it again - and this time, try to read between the lines. ID: 331166 · |
|
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0
|
If you read the first post you'll see what the point is. If you read the white paper that describes the original concepts behind BOINC, one of those (see section 2.1) is to reward credit across diverse projects. If this is in fact something that has value, then all projects should compare their credit rate against the other projects, not just SETI. ID: 331197 · |
|
Saenger Send message Joined: 3 Apr 99 Posts: 2452 Credit: 33,281 RAC: 0
|
Results from my current machine (Seti has just restarted, have to wait yet for the first enhanced to get Credit) Here is finaly the Seti: Enhanced Projekt WU Time [sec] Claimed Granted C/h claim Seti 80950112 32895,56 61,24 61,24 6,70 So even the stock application overclaims here (other Projects average: 5,25) Und zu Jochen: Even in Seti WUs af different angle range have to be measured with different points/credits to be fair. As the most fun with BOINC is the multiple project aspect, it's even more important to get it even. We all just calculate gazillions of Ones and Zeroes, regardless of the aim of the project. These 101010101010 mean something or the other on this and that project, but in principle it's all just the same. So it's not apples and bananas, but imho even Granny Smith with Granny Smith if it's done properly. Gruesse vom Saenger For questions about Boinc look in the BOINC-Wiki ID: 332125 · |
|
Josef W. Segur Send message Joined: 30 Oct 99 Posts: 4504 Credit: 1,414,761 RAC: 0
|
Results from my current machine (Seti has just restarted, have to wait yet for the first enhanced to get Credit) Because you're running Linux, claims based on benchmarks tend to be low. Is the 5.25 average claimed or granted? Joe ID: 332215 · |
|
Saenger Send message Joined: 3 Apr 99 Posts: 2452 Credit: 33,281 RAC: 0
|
Claimed, this is over all: Time [sec] Claimed Granted C/h claim C/h grant 467847,33 681,63 757,17 5,25 5,83 So I get my share by granting, but not 6,7. ID: 332234 · |
©2025 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.