Pages: [1]
[AF>Libristes]Excelsior
BAM!ID: 18403
Joined: 2007-02-02
Posts: 10
Credits: 25,894,565
World-rank: 27,323

2012-04-12 12:17:49

Hi,

I've joined back distributed computation since a few days, and I'm quite worried about BOINC's future.
You'll understand why in the title of this thread, it's about distrRTgen credits. I am amazed by how much credits are given, an it's encouraging people to crunch data for this project at the exclusion of any others. This is what bugs me the most. I don't crave for first place, but I admit the competition is (a little bit) exciting and I even thought about crunching for this project, even though I don't care about what the project is doing.

Here's a comparison of figures made by distrRTgen and Seti@Home :

            Figure               |   distRTgen    |   Seti@Home

    Credits the lasts 60 days    | 12,000,000,000 | 6,000,000,000

       Active computers          |          4,092 |       244,944

Average credits / day / computer |         48,876 |           408


It's not necessary to show more explanation, there is only 2 possibilities :
- Either there is 4,100 monsters crunching for distrRTgen and 245,000 intel 486 crunching for Seti@Home
- Either distrRTgen gives credits a hundred times bigger than Seti@Home

I hope something will be done for the sake of balance because, for it's just encouraging crunching for buggy projects, and I really don't feel this is BOINC's philosophy.

Thanks for reading.

PS : Don't bother to rant against me. I'm not dumb, you can explain me why I'm wrong if you think so.

Guest

2012-04-12 14:10:25

it's your computer, run what you wish to run.
there are people that run just for the credits and others that run for science and others that run both.
as for me, if my computers will run it, I run it. doesn't matter to me if it math, science, etc. I run for everybody, as long as all my computers can run it. and yes it is very hard not to get caught up in the credit thing. just remember to have fun. if it something you want to run and your having fun at it then go for it. just have fun and don't turn it into a job where you will hate it and pull the plug. too many (about 75%) of users that start running projects drop out. "HAVE FUN"
rroonnaalldd
BAM!ID: 15220
Joined: 2006-12-20
Posts: 207
Credits: 77,733,697
World-rank: 12,723

2012-04-12 17:54:35

Point 1. DistrRTgen had installed new hardware and want to test the overall project stability with the Rainbow Resurrection Challenge between 2012-04-07 16:00:00 GMT - 2012-04-10 16:00:00 GMT.
To get more load on the servers, the admins increased the credits for the challenge.

Point 2. Seti is not the centre of the boinc world and only one of 100 other projects. A credit parity between all projects is only the impossible dream of DA.

Point 3. Crunch what ever you want and have fun in doing it.
STE\/E
Tester
BAM!ID: 57534
Joined: 2008-08-27
Posts: 2078
Credits: 1,910,943,363,872
World-rank: 2

2012-04-12 22:19:51


Point 4:

If you want to run with the Big Dogs you have to get on the Porch ... lol
[AF>Libristes]Excelsior
BAM!ID: 18403
Joined: 2007-02-02
Posts: 10
Credits: 25,894,565
World-rank: 27,323

2012-04-13 10:20:56

Thanks for giving your point of view.

Surely, you run what you want to run, the point of BOINC is not to compete or race, but it adds a small fun whatsoever.

Just to make it clear, I selected SETI as reference because it's an old project which can be used as a solid reference, which credit attribution is neither over nor under rated (or so I think).
Of course, there are lots of other projects, but my point wasn't to do a full comparison between each projects (this could be a good idea however).
Maybe I'm a dreamer too, because the credit parity thing is something I would really enjoy.

I only hope that people here don't really give a damn about point 4.

Thanks, and have fun crunching.
zombie67
BAM!ID: 1560
Joined: 2006-06-04
Posts: 606
Credits: 178,101,932,854
World-rank: 45

2012-04-13 19:43:15
last modified: 2012-04-13 19:44:02

AFExcelsior wrote:
Of course, there are lots of other projects, but my point wasn't to do a full comparison between each projects (this could be a good idea however).


It already exists:

http://boincstats.com/stats/project_cpcs.php

AFExcelsior wrote:
Maybe I'm a dreamer too, because the credit parity thing is something I would really enjoy.


It is an impossibility, for many reasons.
Reno, NV
Team: SETI.USA
[AF>Libristes]Excelsior
BAM!ID: 18403
Joined: 2007-02-02
Posts: 10
Credits: 25,894,565
World-rank: 27,323

2012-04-13 20:27:57

zombie67 wrote:
AFExcelsior wrote:
Of course, there are lots of other projects, but my point wasn't to do a full comparison between each projects (this could be a good idea however).


It already exists:

http://boincstats.com/stats/project_cpcs.php

I already peeked at it, yeah.
I hope you realised that this chart cannot be trusted. If you compare Seti and distrRTgen with this chart, it's saying distrRTgen gives 1.185x more credits than seti but it's giving almost 100x more.

zombie67 wrote:
AFExcelsior wrote:
Maybe I'm a dreamer too, because the credit parity thing is something I would really enjoy.


It is an impossibility, for many reasons.

Well, I sure understand, but one can dream
zombie67
BAM!ID: 1560
Joined: 2006-06-04
Posts: 606
Credits: 178,101,932,854
World-rank: 45

2012-04-14 15:34:51

AFExcelsior wrote:
I hope you realised that this chart cannot be trusted. If you compare Seti and distrRTgen with this chart, it's saying distrRTgen gives 1.185x more credits than seti but it's giving almost 100x more.


That depends on which kind of processor (CPU, ATI, nvidia) and which applications (stock, optimized, MB, AP, etc) you are using for your comparison. Which takes us back to the impossibility of cross-project credit parity.
Reno, NV
Team: SETI.USA
Jazzcaster
BAM!ID: 122543
Joined: 2012-04-23
Posts: 1
Credits: 72,011,539
World-rank: 13,451

2012-04-24 20:39:13

From my experience using a the same GPU on SETI as DistrRTGen, the latter gives approximately 20 times the credits as SETI for the "same amount" (time only) of work.
Aleksander Parkitny
 
BAM!ID: 13554
Joined: 2006-11-30
Posts: 6
Credits: 119,788,239
World-rank: 9,441

2012-06-26 13:11:32

Actualli it is better to compare similar projects. Like colltaz Milkyway and DistRTGen

Collatz RAC - 105,650,136 Computers - 8,981 (active hosts) RAC/Host - 11763
DistRTGen RAC - 315,079,364 Computers - 4,299 RAC/Host - 73291
Milkyway RAC - 104,926,951 Computers - 36,488 RAC/Host - 2875

now Collatz mily and Dist are all crunching mainly on GPUs and You can clearly see that milkyway is grantong less than 3000 points per host Collatz gives less than 12000 points per host and dist is giving over 70000 credits per host almost 7 times more than Collatz and almost 30 times more than Milky. it is simply a matter of credit inflation. Maybe other project admins should set their credits as high? that would be fun indeed. I remember the case of FreeHAL and i've earned 60milions of credits in a couple of days but after short notice from Boincstats Tobias reseted the point to normal levels. I think the same should be done with DistRTGen
Pages: [1]

Index :: The Projects :: DistrRTgen credits attributions
Reason: