| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: The Clean Energy Project Forum
Thread: Points calculation |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 18
|
|
| Author |
|
|
crille1006
Advanced Cruncher Germany Joined: Aug 15, 2007 Post Count: 79 Status: Offline Project Badges: 
|
E000002_ 752A_ 00001t00m_ 0-- Valid 06.12.08 16:00:03 08.12.08 17:10:02 8.30 83.2 / 83.2
----------------------------------------E000002_ 752A_ 00001t00m_ 1-- Valid 06.12.08 15:55:25 08.12.08 07:59:29 28.10 326.3 / 83.2 Hi! I don't think, that one computer needed 8.30 hrs and another 28.10. Is it possible, that one task was aborted but both are valid? ---------------------------------------- [Edit 1 times, last edit by crille1006 at Dec 8, 2008 8:18:30 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
There is nothing exceptional about one computer being 3-4 times faster than another.
The claimed credit disparity is a concern. Possibly the high claiming computer went into a power-saving mode and thus crunched it at a slower speed, which still reporting the normal benchmark result. As power saving features become more common, this is becoming a significant problem for us. Aborted results are marked "Error". |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
The exit code for results to be considered for validation is 0. That said, we've seen something similar to RICE. It's not clear [to me] and have asked in back room to have a look v.v. a specific log result error, where the [short] results still validated.
----------------------------------------It's possible a result has useful information, but as the project scientist wrote yesterday or the day before, they are evaluating the first production results further (for optimization, improvement and sizing).
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Nutritious Rice for the World works differently. Because the application simply does as many structure predictions as it can in the time-limit, if the program exits early, then it is exactly the same as if it had done the same number of predictions really slowly. The (however many) predictions that succeeded can then be validated normally.
The Clean Energy Project does not work in this way. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
We're surely going learn if astrolabs or crille1006's machine went into any power saving mode. The unexplained log error reports bug me more.
----------------------------------------
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
We're surely going learn if astrolabs or crille1006's machine went into any power saving mode. While I am not certain about the power saving on that computer, I am certain that the setting has not been changed since June and since the computer is set to run at 100% the significant difference was unexpected. |
||
|
|
crille1006
Advanced Cruncher Germany Joined: Aug 15, 2007 Post Count: 79 Status: Offline Project Badges:
|
My computer is the fast one with 8.30 hrs runtime. I noticed that the workunit progress was after 6hrs runtime at 40%. Only 2 hrs later the wu was finished and it seems to me, that the progress status jumps to 100%...
---------------------------------------- |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Guys....
You are talking about two different work units. crille1006's quorum partner is the one with the energy saving problem. Nothing else can explain an anomaly that large. The difference in astrolab's quorum is considerably less. astrolab, does this computer consistently overclaim? Some computers do. Or, perhaps you are affected by a high pagefault count. That's something you can check, at least. |
||
|
|
|