Advanced search

Message boards : Webpage and Application Code Discussion : Excessive credit?

1 · 2 · Next
Author Message
adrianxw
Send message
Joined: 31 May 17
Posts: 12
Combined Credit: 5,728,418
DNA@Home: 0
SubsetSum@Home: 0
Wildlife@Home: 5,728,418
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 0

  
Message 7127 - Posted: 9 Jun 2017, 8:18:38 UTC

I've only been a cruncher here for a few days, yet, my credit total now exceeds 500,000, more than other projects I have been crunching for years. The quota for the project is not high.
____________
Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.

adrianxw
Send message
Joined: 31 May 17
Posts: 12
Combined Credit: 5,728,418
DNA@Home: 0
SubsetSum@Home: 0
Wildlife@Home: 5,728,418
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 0

  
Message 7133 - Posted: 10 Jun 2017, 8:20:33 UTC - in response to Message 7127.

... and a day later, exceeds 700,000.
____________
Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.

Profile Gibson Praise
Avatar
Send message
Joined: 5 Aug 16
Posts: 52
Combined Credit: 36,447,624
DNA@Home: 0
SubsetSum@Home: 73,256
Wildlife@Home: 36,374,368
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 768
Images Observed: 23,296

        
Message 7135 - Posted: 10 Jun 2017, 20:22:51 UTC
Last modified: 10 Jun 2017, 20:25:50 UTC

There was an earlier discussion ( This thread ) back in February. At that time I got this for W@H:

Notes: measurements in terms of # seconds to produce a BOINC point (seconds for unit divided by points granted).

System Spec: Xeon E5-2650 @2.1 GHz running Win10

EXACT: running in between 17 - 21 sec/pt

recent returns have me at 6.5 - 7 sec/pt (due to code and/or compilation improvements I believe).
____________

Travis Desell
Volunteer moderator
Project administrator
Project developer
Project scientist
Send message
Joined: 16 Jan 12
Posts: 1799
Combined Credit: 6,133,999
DNA@Home: 293,563
SubsetSum@Home: 349,212
Wildlife@Home: 5,491,224
Wildlife@Home Watched: 212,926s
Wildlife@Home Events: 51
Climate Tweets: 21
Images Observed: 718

              
Message 7150 - Posted: 20 Jun 2017, 20:02:28 UTC - in response to Message 7135.

If credit is too high I can tone it down a bit. Thoughts?

Profile JumpinJohnny
Avatar
Send message
Joined: 24 Sep 13
Posts: 235
Combined Credit: 7,657,258
DNA@Home: 192,548
SubsetSum@Home: 201,740
Wildlife@Home: 7,262,971
Wildlife@Home Watched: 55,997,833s
Wildlife@Home Events: 15,584
Climate Tweets: 321
Images Observed: 351

              
Message 7167 - Posted: 25 Jun 2017, 4:45:34 UTC - in response to Message 7150.

My highest ten days with CSG have all been between May 29 and June 24.
I have reached or am close to several personal goals.
My high credit total (for me), has been the result of sticking with the variety of trials and WU experiments since last December. It's been well worth it.
I would be somewhat dishonest to say I don't like the current credit payoffs. I would also be dishonest to say that they are not on the high side for a CPU WU.

adrianxw
Send message
Joined: 31 May 17
Posts: 12
Combined Credit: 5,728,418
DNA@Home: 0
SubsetSum@Home: 0
Wildlife@Home: 5,728,418
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 0

  
Message 7168 - Posted: 25 Jun 2017, 9:29:09 UTC
Last modified: 25 Jun 2017, 9:38:52 UTC

My credit is over a million now, in 2 weeks crunching, at regular project priority. Compared with say Rosetta, another CPU based project, this has almost 1.3 million, but has been crunching here at above normal priority, since 2005.

Of CPU based projects, all but 1 have a recent average under 5000, the last has an average of 7000. The recent average here is just short of 30000

I maintain the credit here is too high.
____________
Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.

DJStarfox
Send message
Joined: 23 Jan 15
Posts: 11
Combined Credit: 853,774
DNA@Home: 19,182
SubsetSum@Home: 0
Wildlife@Home: 834,592
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 0

    
Message 7356 - Posted: 13 Nov 2017, 14:36:34 UTC

I agree that RAC is too high compared to other projects. Typically, projects benchmark their workunits against the credit granted by a SETI@Home workunit.

Profile Coleslaw
Avatar
Send message
Joined: 22 Jun 12
Posts: 22
Combined Credit: 26,175,516
DNA@Home: 569,863
SubsetSum@Home: 1,698,846
Wildlife@Home: 23,906,806
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 15
Images Observed: 0

        
Message 7364 - Posted: 24 Nov 2017, 17:28:47 UTC - in response to Message 7356.

I agree that RAC is too high compared to other projects. Typically, projects benchmark their workunits against the credit granted by a SETI@Home workunit.


Wait what? What projects have actually done that? SETI uses the Credit New credit system which is all over the place at most projects that try it. If a benchmark is set for comparison, the project admins typically do it based on the community as a whole. Not based on SETI. So, what they typically will do when setting theirs is either used fix credit per work unit or they will base it off of what a specific processor's performance is. SETI is also considered by the general community to under credit and would be a horrible idea as a basis.
____________

UBT - Timbo
Send message
Joined: 31 Dec 13
Posts: 1
Combined Credit: 2,483,812
DNA@Home: 45,472
SubsetSum@Home: 169,611
Wildlife@Home: 2,268,729
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 0

      
Message 7365 - Posted: 24 Nov 2017, 20:39:51 UTC

Hi all

Although, in an idealistic world, the credits issued by a project shouldn't matter, in reality, they do to a lot of people.

As such, I think the projects should set the credit level to be commensurate with how quickly they want to see results from the tasks that they make available.

A higher than average credit earning capability will (usually) attract more crunchers to the project and hence the project scientists will get back more meaningful results in a shorter space of time.

So, I have no issue with any project setting too high or too low a level of credit, for any amount of validated tasks, uploaded by crunchers.

Of course, the credit offered can be varied by each project depending on the level of difficulty each type of task presents....

So, where's the problem? Some projects offer lower than average credits per task (SETI, Sztaki) and others are higher.

As long as everyone crunching the same type of tasks on the same project using a similar CPU, gets the same level of credit, then what the actual credit ratio is (compared to CPU or GPU time), should be down to the project.

regards
Tim

DJStarfox
Send message
Joined: 23 Jan 15
Posts: 11
Combined Credit: 853,774
DNA@Home: 19,182
SubsetSum@Home: 0
Wildlife@Home: 834,592
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 0

    
Message 7377 - Posted: 12 Dec 2017, 14:31:28 UTC

The credit system only works if the baseline metric is the same (or similar) as all other projects. When one project grants significantly less or more credit than other projects, it creates an unfairness that encourages other projects to defect. Granted credit would then vary wildly between projects for the same CPU/GPU. The end result is the invalidation of credit as a measurement of user contribution. (I.e., credits become meaningless without a standard metric/unit to compare against.)

In regards to my SETI comment earlier, back on the old credit system, SETI had a numeric scale for measuring work done that could be translated to other projects. I'm not familiar with how it's done now. Perhaps an expert at another project would be helpful for a consult? Or one of these technical guys:
http://boinc.berkeley.edu/trac/wiki/ProjectGovernance

adrianxw
Send message
Joined: 31 May 17
Posts: 12
Combined Credit: 5,728,418
DNA@Home: 0
SubsetSum@Home: 0
Wildlife@Home: 5,728,418
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 0

  
Message 7378 - Posted: 13 Dec 2017, 13:15:32 UTC

I've tried to get this fixed, but he doesn't seem to want too. At the same Resource share setting, I get 5 times the credit here as I do at other CPU projects, it is by some way, my highest credit awarder, 5-6 times the next highest. I gave up bothering, credit is of no use anyway.
____________
Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.

Travis Desell
Volunteer moderator
Project administrator
Project developer
Project scientist
Send message
Joined: 16 Jan 12
Posts: 1799
Combined Credit: 6,133,999
DNA@Home: 293,563
SubsetSum@Home: 349,212
Wildlife@Home: 5,491,224
Wildlife@Home Watched: 212,926s
Wildlife@Home Events: 51
Climate Tweets: 21
Images Observed: 718

              
Message 7379 - Posted: 14 Dec 2017, 15:19:53 UTC - in response to Message 7378.

I'm waiting for a couple batches of runs to complete, and then have some new code to test so I'll probably keep credit at the current rate for a bit.

After the next phase of testing and getting things stable I will most likely start to lower it to get it more in line with other projects, since it seems like its significantly higher.

adrianxw
Send message
Joined: 31 May 17
Posts: 12
Combined Credit: 5,728,418
DNA@Home: 0
SubsetSum@Home: 0
Wildlife@Home: 5,728,418
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 0

  
Message 7380 - Posted: 14 Dec 2017, 20:20:00 UTC
Last modified: 14 Dec 2017, 20:23:43 UTC

>>> couple batches

We've been saying this since June.....
____________
Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.

Profile Conan
Avatar
Send message
Joined: 13 Apr 12
Posts: 143
Combined Credit: 32,980,985
DNA@Home: 399,792
SubsetSum@Home: 1,448,876
Wildlife@Home: 31,132,317
Wildlife@Home Watched: 70,910s
Wildlife@Home Events: 0
Climate Tweets: 396
Images Observed: 0

          
Message 7381 - Posted: 15 Dec 2017, 3:44:37 UTC - in response to Message 7380.

>>> couple batches

We've been saying this since June.....


Sarcasm and being ungrateful seem to go hand in hand these days, which is a pity.

If you did the amount of work that Travis does across multiple projects (not just CSG) and teaching classes, then you should be amazed at the lack of problems that arise across his projects.

Credit (or points) are a minor thing compared to getting a project running smoothly and staying running with all the work behind the scenes as well.

The projects work units run my computers very hard and hot, so a credit (or point) bonus for running them is fine by me.

Being sarcastic to the admin of a well run project (and responsive as well), will not win friends, being logical and polite just might.

Of course we have all heard it before that if we don't like what a project is doing then we are free to leave, and that course of action is open to all.

I am happy with the current status, probably because I run so many other projects that pay me bugger-all even compared to SETI, which is not the greatest payer of points, so I believe that some higher paying projects help to even my credit out overall.

Please be patient as there is only one Travis and we don't want to wear him out.

Conan
____________

adrianxw
Send message
Joined: 31 May 17
Posts: 12
Combined Credit: 5,728,418
DNA@Home: 0
SubsetSum@Home: 0
Wildlife@Home: 5,728,418
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 0

  
Message 7382 - Posted: 15 Dec 2017, 15:10:28 UTC

I'm a little puzzled by this.

>>> Sarcasm and being ungrateful

I see no sarcasm, that is a simple statement of fact. He has said it will be fixed, and it has not been. You simply need to look at the dates of the posts in this thread to see that. He acknowledges the basically unwritten rule that projects aim to keep credit inline, he is not doing that.

Viz ungrateful, that puzzles me even more. We are doing his work, who do you think should be grateful?

>>> If you did the amount of work

I crunch many projects, under several names and have been doing so since the early predecessors of BOINC in the last century.
____________
Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.

Travis Desell
Volunteer moderator
Project administrator
Project developer
Project scientist
Send message
Joined: 16 Jan 12
Posts: 1799
Combined Credit: 6,133,999
DNA@Home: 293,563
SubsetSum@Home: 349,212
Wildlife@Home: 5,491,224
Wildlife@Home Watched: 212,926s
Wildlife@Home Events: 51
Climate Tweets: 21
Images Observed: 718

              
Message 7383 - Posted: 15 Dec 2017, 22:35:01 UTC - in response to Message 7382.

Sorry I do respond a bit slow at times. One of my issues is the next sets of workunits (after two more runs) are going to be on the CIFAR-10 dataset as opposed to the MNIST dataset. These will be approximately 3x as large which may introduce some weird memory issues which we were seeing before. I kind of like giving people some bonus credits when I know things will be a bit buggy as a bonus for putting up with the development process.

But that being said, the credit rate probably does need to be dropped a bit and I think I'll do this gradually so I don't do anything to drastic.

It won't happen for a few days though. The computer science department at UND is physically moving to a different building and this all needs to be completed by the coming monday. Which also happens to be the day before grades are due (thanks administrators!). :P So it's been a pretty hectic semester to say the least.

adrianxw
Send message
Joined: 31 May 17
Posts: 12
Combined Credit: 5,728,418
DNA@Home: 0
SubsetSum@Home: 0
Wildlife@Home: 5,728,418
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 0

  
Message 7384 - Posted: 17 Dec 2017, 9:08:27 UTC
Last modified: 17 Dec 2017, 9:08:55 UTC

Of CPU projects on here at base priority, (ie. 100), their RAC varies between 1500 and 8000, of course, work unit availability needs to be factored in there. Here. also at 100 priority, my RAC right now is 25,244.

As you said...

>>> the credit rate probably does need to be dropped a bit
____________
Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.

DJStarfox
Send message
Joined: 23 Jan 15
Posts: 11
Combined Credit: 853,774
DNA@Home: 19,182
SubsetSum@Home: 0
Wildlife@Home: 834,592
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 0

    
Message 7385 - Posted: 18 Dec 2017, 14:51:12 UTC - in response to Message 7383.

Travis:

I, for one, am grateful to have a project administrator paying attention and making needed changes to this active project. That sounds like a lot of crazy work for the end of a semester. I'm sure working together, we will continue to improve the science and technicals.

Hope you have a happy holiday season.

Profile Philip
Send message
Joined: 7 May 14
Posts: 1
Combined Credit: 1,292,167
DNA@Home: 17,882
SubsetSum@Home: 101,258
Wildlife@Home: 1,173,027
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 1,391

        
Message 7386 - Posted: 20 Dec 2017, 9:46:53 UTC

I vote to leave credits as is

Philip
____________

Jozef J
Send message
Joined: 19 Apr 14
Posts: 15
Combined Credit: 21,881,667
DNA@Home: 1,006,826
SubsetSum@Home: 115,384
Wildlife@Home: 20,759,457
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 3,451
Images Observed: 37,019

          
Message 7387 - Posted: 20 Dec 2017, 16:27:08 UTC

I vote to leave credits as it.
J.J.

1 · 2 · Next
Post to thread

Message boards : Webpage and Application Code Discussion : Excessive credit?