Welcome to MilkyWay@home

credit comparison to other projects

Message boards : Number crunching : credit comparison to other projects
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 3 · 4 · 5 · 6 · 7 · 8 · 9 . . . 15 · Next

AuthorMessage
Profile caspr
Avatar

Send message
Joined: 22 Mar 08
Posts: 90
Credit: 501,728
RAC: 0
Message 7347 - Posted: 3 Dec 2008, 18:00:34 UTC - in response to Message 7343.  
Last modified: 3 Dec 2008, 18:18:04 UTC

I'm not gonna get all philosophical on this but I WAS getting about 9k here (milsop's app.) and now less than 1.5k per day.( new opp app) I'd just like to see a reasonable adjustment.


again, how does that compare to other projects? also if you were using the optimized application on older machines chances are it's not bumping up to near the credit limit anymore, so you're seeing more of a penalty.



I'm running an '03server 3.0 xeon DDR2-400 Ecc Reg's So its a Quad... kinda.

EDIT: Is this about bumping the client or fair credits per cpu?
EDITx2 Sorry I'm not running an i7!
A clear conscience is usually the sign of a bad memory



ID: 7347 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile AstralWalker

Send message
Joined: 19 Feb 08
Posts: 14
Credit: 975,511
RAC: 274
Message 7350 - Posted: 3 Dec 2008, 18:49:23 UTC

Maybe we need a separate BOINC project to determine how credit should be awarded.
ID: 7350 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Arion
Avatar

Send message
Joined: 10 Aug 08
Posts: 218
Credit: 41,846,854
RAC: 0
Message 7352 - Posted: 3 Dec 2008, 19:12:59 UTC - in response to Message 7335.  
Last modified: 3 Dec 2008, 19:31:45 UTC

Arion, your calculations surprise me, since for me Einstein gives better credit atm...


Monday I think it was I upgraded to the Power App (in testing) .07, I was using .05 and it seems it has slowed einstein down a bit here in comparison to other workunits I have before the switch over. I'm not sure if I'm hitting the top of a trough. It was 2 workunits that spanned the same amount of time that the ones from MW processed.If I remember right last time I checked it was in the area of .52 or something.

Otherwise were my calculations based on the correct formula?

Edit - I just ran a quick check on pairs of previous ones before I switched over to the latest release on the power app and the average was running around .50 - .52
ID: 7352 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Travis
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 30 Aug 07
Posts: 2046
Credit: 26,480
RAC: 0
Message 7356 - Posted: 3 Dec 2008, 19:23:23 UTC - in response to Message 7347.  

I'm not gonna get all philosophical on this but I WAS getting about 9k here (milsop's app.) and now less than 1.5k per day.( new opp app) I'd just like to see a reasonable adjustment.


again, how does that compare to other projects? also if you were using the optimized application on older machines chances are it's not bumping up to near the credit limit anymore, so you're seeing more of a penalty.



I'm running an '03server 3.0 xeon DDR2-400 Ecc Reg's So its a Quad... kinda.

EDIT: Is this about bumping the client or fair credits per cpu?
EDITx2 Sorry I'm not running an i7!


and how do the credits awarded from MW compare to other projects?
ID: 7356 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Brian Silvers

Send message
Joined: 21 Aug 08
Posts: 625
Credit: 558,425
RAC: 0
Message 7358 - Posted: 3 Dec 2008, 19:33:45 UTC - in response to Message 7350.  

Maybe we need a separate BOINC project to determine how credit should be awarded.


Credit@Home?

Look out for users named "AIG", "FannieMae", or "Citi"...
ID: 7358 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Arion
Avatar

Send message
Joined: 10 Aug 08
Posts: 218
Credit: 41,846,854
RAC: 0
Message 7360 - Posted: 3 Dec 2008, 19:52:07 UTC - in response to Message 7356.  

and how do the credits awarded from MW compare to other projects?



Based on my calculations Travis (only on one system against einstein - NOT Boinc Wide) With Milksap's app I was capped at 108 credits per hour. On Einstein a workunit there takes average 6 1/2 to 8 hours and credit awarded is 222.43.

Now with the new app here it is more reasonable in that I'm getting about 40+ credits per hour. (ballpark). I realize that my figures are only for 1 of my systems - AMD 64 X2 6400 stock and it is not maybe the norm with the majority of other users. I have 3 slower systems that are taking much longer to process at einstein and here.. If I use this as a general example then in relation to Einstein you are a little high BUT not to any kind of extreme. Based on the comparison chart that is being discussed I say you are just above some and a lot lower than others. Is this a fair amount of credits? I don't have any problems but you ultimately make the decision.

To put it in another perspective, I don't think at this point you have to worry about DA paying you another visit. <cough, cough>



ID: 7360 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Arion
Avatar

Send message
Joined: 10 Aug 08
Posts: 218
Credit: 41,846,854
RAC: 0
Message 7361 - Posted: 3 Dec 2008, 19:54:09 UTC - in response to Message 7358.  

Maybe we need a separate BOINC project to determine how credit should be awarded.


Credit@Home?

Look out for users named "AIG", "FannieMae", or "Citi"...



Should we look for a bailout if our credits dry up? <Smile>

ID: 7361 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Cori
Avatar

Send message
Joined: 27 Aug 07
Posts: 647
Credit: 27,592,547
RAC: 0
Message 7364 - Posted: 3 Dec 2008, 20:08:15 UTC - in response to Message 7310.  

... nm_stripe86 does more work than nm_stripe79 and nm_stripe82, but it awards more credit based on the extra work. You should be getting more credit for the longer WUs.

Hm, I don't understand it: my nm_stripe86 get less credits: only 21,18 per WU but the nm_stripe79 and nm_stripe82 do get ~39 credits. *scratches head*
Lovely greetings, Cori
ID: 7364 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John Clark

Send message
Joined: 4 Oct 08
Posts: 1734
Credit: 64,228,409
RAC: 0
Message 7365 - Posted: 3 Dec 2008, 20:20:09 UTC - in response to Message 7356.  
Last modified: 3 Dec 2008, 20:20:55 UTC

and how do the credits awarded from MW compare to other projects?


I have just been doing some SETI, to build the RAC slightly, and the claimed and given credit per hour is similar to MW (based on a short 20 Wu run.
ID: 7365 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Travis
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 30 Aug 07
Posts: 2046
Credit: 26,480
RAC: 0
Message 7370 - Posted: 3 Dec 2008, 21:15:07 UTC - in response to Message 7364.  

... nm_stripe86 does more work than nm_stripe79 and nm_stripe82, but it awards more credit based on the extra work. You should be getting more credit for the longer WUs.

Hm, I don't understand it: my nm_stripe86 get less credits: only 21,18 per WU but the nm_stripe79 and nm_stripe82 do get ~39 credits. *scratches head*


really? stripe_86 should be worth more credits than 79 and 82... the database says stripe 86 should be generating ~42 credits while 79 and 82 are ~39 credits
ID: 7370 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Brian Silvers

Send message
Joined: 21 Aug 08
Posts: 625
Credit: 558,425
RAC: 0
Message 7371 - Posted: 3 Dec 2008, 21:22:08 UTC - in response to Message 7361.  

Maybe we need a separate BOINC project to determine how credit should be awarded.


Credit@Home?

Look out for users named "AIG", "FannieMae", or "Citi"...



Should we look for a bailout if our credits dry up? <Smile>



No, it's if their credits dry up. In that situation, Credit@Home will issue them "Credit Bonds" (workunits) while the rest of us note that we keep getting told "There was credit (work) available for other platforms"
ID: 7371 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Cori
Avatar

Send message
Joined: 27 Aug 07
Posts: 647
Credit: 27,592,547
RAC: 0
Message 7372 - Posted: 3 Dec 2008, 21:22:16 UTC - in response to Message 7370.  

... nm_stripe86 does more work than nm_stripe79 and nm_stripe82, but it awards more credit based on the extra work. You should be getting more credit for the longer WUs.

Hm, I don't understand it: my nm_stripe86 get less credits: only 21,18 per WU but the nm_stripe79 and nm_stripe82 do get ~39 credits. *scratches head*


really? stripe_86 should be worth more credits than 79 and 82... the database says stripe 86 should be generating ~42 credits while 79 and 82 are ~39 credits

Hmmm... I have had another look at my results and found a nm_stripe86 with 42 credits but the newer ones get only the half of the credits:

"old" nm_stripe86: http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=57002953

"new" nm_stripe86: http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=57010436
Lovely greetings, Cori
ID: 7372 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Labbie
Avatar

Send message
Joined: 29 Aug 07
Posts: 327
Credit: 116,463,193
RAC: 0
Message 7373 - Posted: 3 Dec 2008, 21:26:53 UTC - in response to Message 7372.  

... nm_stripe86 does more work than nm_stripe79 and nm_stripe82, but it awards more credit based on the extra work. You should be getting more credit for the longer WUs.

Hm, I don't understand it: my nm_stripe86 get less credits: only 21,18 per WU but the nm_stripe79 and nm_stripe82 do get ~39 credits. *scratches head*


really? stripe_86 should be worth more credits than 79 and 82... the database says stripe 86 should be generating ~42 credits while 79 and 82 are ~39 credits

Hmmm... I have had another look at my results and found a nm_stripe86 with 42 credits but the newer ones get only the half of the credits:

"old" nm_stripe86: http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=57002953

"new" nm_stripe86: http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=57010436


I'm seeing this too.

57212793


Calm Chaos Forum...Join Calm Chaos Now
ID: 7373 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Travis
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 30 Aug 07
Posts: 2046
Credit: 26,480
RAC: 0
Message 7374 - Posted: 3 Dec 2008, 21:30:47 UTC - in response to Message 7373.  

... nm_stripe86 does more work than nm_stripe79 and nm_stripe82, but it awards more credit based on the extra work. You should be getting more credit for the longer WUs.

Hm, I don't understand it: my nm_stripe86 get less credits: only 21,18 per WU but the nm_stripe79 and nm_stripe82 do get ~39 credits. *scratches head*


really? stripe_86 should be worth more credits than 79 and 82... the database says stripe 86 should be generating ~42 credits while 79 and 82 are ~39 credits

Hmmm... I have had another look at my results and found a nm_stripe86 with 42 credits but the newer ones get only the half of the credits:

"old" nm_stripe86: http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=57002953

"new" nm_stripe86: http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=57010436


I'm seeing this too.

57212793


and these are still running longer than the other ones?
ID: 7374 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Cori
Avatar

Send message
Joined: 27 Aug 07
Posts: 647
Credit: 27,592,547
RAC: 0
Message 7375 - Posted: 3 Dec 2008, 21:38:38 UTC - in response to Message 7374.  

and these are still running longer than the other ones?

The one with only 21 credits was even a bit longer than the one with 42 credits (see my links). ;)
Lovely greetings, Cori
ID: 7375 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Travis
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 30 Aug 07
Posts: 2046
Credit: 26,480
RAC: 0
Message 7376 - Posted: 3 Dec 2008, 21:45:35 UTC - in response to Message 7373.  

... nm_stripe86 does more work than nm_stripe79 and nm_stripe82, but it awards more credit based on the extra work. You should be getting more credit for the longer WUs.

Hm, I don't understand it: my nm_stripe86 get less credits: only 21,18 per WU but the nm_stripe79 and nm_stripe82 do get ~39 credits. *scratches head*


really? stripe_86 should be worth more credits than 79 and 82... the database says stripe 86 should be generating ~42 credits while 79 and 82 are ~39 credits

Hmmm... I have had another look at my results and found a nm_stripe86 with 42 credits but the newer ones get only the half of the credits:

"old" nm_stripe86: http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=57002953

"new" nm_stripe86: http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=57010436


I'm seeing this too.

57212793



I actually just checked out the server, in last nights mess the parameter files for stripe86 got cut down in accuracy by half (which explains why they're giving about half less credits). newly generated stripe86 wus should be giving the right credit (and take the typical time).

the ones that were giving less credit should crunch in about half the time, so the crunch time to credit ratio should still be the same.
ID: 7376 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Arion
Avatar

Send message
Joined: 10 Aug 08
Posts: 218
Credit: 41,846,854
RAC: 0
Message 7377 - Posted: 3 Dec 2008, 21:52:33 UTC - in response to Message 7374.  
Last modified: 3 Dec 2008, 21:58:07 UTC

and these are still running longer than the other ones?


I'm starting to see the same thing here.

NM-Strip86-r2 - 2414.85 - 12.57 - 42.33
NM-Strip86-r2 - 2409.74 - 12.54 - 21.18

Description ---/cpu Time-Claimed-Awarded

Travis, will I have to allow computers to be shown if you need to look at them or can you look at my work history without a public flag?

computer ID is #36995



79's and 82's are getting 39.84....
ID: 7377 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Cori
Avatar

Send message
Joined: 27 Aug 07
Posts: 647
Credit: 27,592,547
RAC: 0
Message 7378 - Posted: 3 Dec 2008, 21:52:54 UTC - in response to Message 7376.  

I actually just checked out the server, in last nights mess the parameter files for stripe86 got cut down in accuracy by half (which explains why they're giving about half less credits). newly generated stripe86 wus should be giving the right credit (and take the typical time).

the ones that were giving less credit should crunch in about half the time, so the crunch time to credit ratio should still be the same.

Ok, but at least once there must have been a hiccup:
"old" nm_stripe86: http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=57002953 CPU time 3551.474 = Granted credit 42.33592

"new" nm_stripe86: http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=57010436 CPU time 3741.538 = Granted credit 21.180365

I have seen a newer WU now, where the crunch time is indeed cut in about half time so there the reduced credits do match:
http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=57012280 CPU time 1765.394 = Granted credit 21.180365

Ok, it seems it was only one WU affected with me, but I tell ya I really got confused at first!! *grin*
Lovely greetings, Cori
ID: 7378 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Labbie
Avatar

Send message
Joined: 29 Aug 07
Posts: 327
Credit: 116,463,193
RAC: 0
Message 7380 - Posted: 3 Dec 2008, 22:02:07 UTC - in response to Message 7374.  


I'm seeing this too.

57212793


and these are still running longer than the other ones?


Looking at mine, they are running about the same CPU time, ~2500 seconds.

Here is a 42 credit one which ran 2512.03 seconds, the one linked above was a 21 credit WU that ran for 2516.56 seconds.

Calm Chaos Forum...Join Calm Chaos Now
ID: 7380 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Travis
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 30 Aug 07
Posts: 2046
Credit: 26,480
RAC: 0
Message 7383 - Posted: 3 Dec 2008, 22:23:44 UTC - in response to Message 7380.  


I'm seeing this too.

57212793


and these are still running longer than the other ones?


Looking at mine, they are running about the same CPU time, ~2500 seconds.

Here is a 42 credit one which ran 2512.03 seconds, the one linked above was a 21 credit WU that ran for 2516.56 seconds.


Actually i just saw the problem why they're running the same ;( While the astronomy_parameters file the assimilator was reading was changed, the sticky one in the download directory wasn't. I don't know if there's too much i can do about the WUs already generated with less credit, but the new ones should award the correct amount.
ID: 7383 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Previous · 1 . . . 3 · 4 · 5 · 6 · 7 · 8 · 9 . . . 15 · Next

Message boards : Number crunching : credit comparison to other projects

©2024 Astroinformatics Group