Welcome to MilkyWay@home

Credit Calculations.

Message boards : Number crunching : Credit Calculations.
Message board moderation

To post messages, you must log in.

1 · 2 · 3 · 4 . . . 9 · Next

AuthorMessage
John McLeod VII
Avatar

Send message
Joined: 27 Aug 07
Posts: 85
Credit: 405,705
RAC: 0
Message 4272 - Posted: 21 Jul 2008, 0:36:51 UTC

Having approximately the same credits per hour for a particular host across projects is what is SUPPOSED to happen in BOINC. If people are tired of hearing about this, then please ask the project administrators to get the calculation right rather than being proud of running off people that keep making that claim.

Projects that generate grossly too high credit claims should be removed from the cross project stats sites as this is just a way of stupidly inflating your cross project credit scores.

The long tasks really should be granting about 50 CS per task. Granting more is anti-social of the project administrators. The project administrators should look at what other projects are granting per hour for a particular host and attempt to match that here.


BOINC WIKI
ID: 4272 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
voltron
Avatar

Send message
Joined: 30 Mar 08
Posts: 50
Credit: 11,593,755
RAC: 0
Message 4273 - Posted: 21 Jul 2008, 1:07:35 UTC

Care for some cheese with that?

Your selection of projects should match your level of comfort and organization.

For most of us it's "run it and gun it".

Voltron
ID: 4273 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Jayargh
Avatar

Send message
Joined: 8 Oct 07
Posts: 289
Credit: 3,690,838
RAC: 0
Message 4274 - Posted: 21 Jul 2008, 1:12:08 UTC - in response to Message 4272.  
Last modified: 21 Jul 2008, 2:08:32 UTC

Having approximately the same credits per hour for a particular host across projects is what is SUPPOSED to happen in BOINC. If people are tired of hearing about this, then please ask the project administrators to get the calculation right rather than being proud of running off people that keep making that claim.

Projects that generate grossly too high credit claims should be removed from the cross project stats sites as this is just a way of stupidly inflating your cross project credit scores.

The long tasks really should be granting about 50 CS per task. Granting more is anti-social of the project administrators. The project administrators should look at what other projects are granting per hour for a particular host and attempt to match that here.



Participants who continually badger projects on this subject aka "Credit Police" should be deleted from the cross project stats IMHO.

Any stats sites that attempt to modify xml stats will be boycotted and thus loose their user base and advertising revenue thus nullifying this superior Boinc/Seti attitude that they are in contol...."Its the users stupid" to paraphrase which have control sorry ;) I will 'volunteer to lead the boycott'...now banish me.

Its nice to know that someone thinks they can regulate 50,80,100 projects this way ;)

If I am not mistaken doesn't the Milkyway Project use optimized apps? Thereby conforming to the side rule governing such apps? Thereby conforming to the prescribed Gestapo credit for the original un-optimized app?

Seti opened the credit Pandora's box with their allowed use and abuse of optimized applications....and now they attempt this strong arm approach again to squash it because it detracts from their user base?

Oh gosh darn I just realized Seti just changed their credit awards and philosophy ...guess this means all the good little projeckts must fall in line to das furor....or else!


Yes you do detect bitterness on my behalf based on years of cross project parity shenanigans....,political at best .....controlling at worst.

Yes JMVII they should evaluate against Cosmology as Milkyway is way too low in comparasin, and Milkyway has said optimized apps exonerating it.Please post this over there as I am ready fer ya ;)

They send one minion after another don't they? At least this one didn't try to hijack a thread.
ID: 4274 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John McLeod VII
Avatar

Send message
Joined: 27 Aug 07
Posts: 85
Credit: 405,705
RAC: 0
Message 4275 - Posted: 21 Jul 2008, 2:17:05 UTC

Please keep personalities out of it.

The design of BOINC was to have approximate cross project parity in credits. Any project that is grossly out of line intentionally is intentionally breaking a design goal of BOINC, and has been a goal of BOINC since the very beginning.

The best method for credit granting is FLOPS counting. However, that is not possible in some cases.

Second best is fixed, but this only works if the tasks are all pretty much the same length.

Third best is to have 2 computers do the same task as this prevents cheating on both the science and on credits. Single tasks per WU only works if there is a quick validation from the answer back to the original question. And in most projects, getting quickly from the answer back to the question is not possible, so 2 tasks is a requirement to make certain the science is done. Then to prevent credit cheating (which is trivial in the single task / WU case and in the case where the project uses the average instead of dropping the high request) drop the high of the two (and if there are three or more, drop the low), average the rest.

What doesn't work well is fixed credits where the tasks vary greatly in length, nor does automatically granting the credit request of a single task unverified against another task.


BOINC WIKI
ID: 4275 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Jayargh
Avatar

Send message
Joined: 8 Oct 07
Posts: 289
Credit: 3,690,838
RAC: 0
Message 4276 - Posted: 21 Jul 2008, 2:24:32 UTC
Last modified: 21 Jul 2008, 2:37:22 UTC

Lets not forget that Seti started the 3rd party optimized app which is inherently unfair to the average user as he has to be aware of,capable of, and d/l said app.....that way they can say a few thousand users run the standard app and we can help the "informed" users inflate their credits thus getting BIG crunchers to stay with Seti and not look elsewhere.....what a way to run a business eh JM Vll?

If BOINC admin is really serious about cleaning up this mess then they will stop Seti from allowing 3rd party apps and take them internal to allow all users to benefit. 1st clean up your mess closer to home before attacking other projects credit awards.Otherwise your points become meaningless....

Here everyone has a chance to be optimized based on their architecture/OS and controlled by the project....much fairer and scientifically sound....but of course the facts and my rant may be ringing your ears by now so I will desist at the moment as I have said what is needed on this subject and hope by now you get my point.
ID: 4276 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Jayargh
Avatar

Send message
Joined: 8 Oct 07
Posts: 289
Credit: 3,690,838
RAC: 0
Message 4277 - Posted: 21 Jul 2008, 2:27:35 UTC - in response to Message 4275.  
Last modified: 21 Jul 2008, 3:09:17 UTC

Please keep personalities out of it.




Only if you stop threatening projects with xml stat manipulation or non-inclusion :)

I am aware (as well as you should be) that this is being done at the highest levels of BOINC to various project administrations.Aren't you a contributor to BOINC programming and hence admin or privy to?

Why is this situation with unfair Seti optimized allowed to continue while BOINC is pushing for cross project parity? This is what needs answering.

Come back and talk cross project parity when that situation with Seti is fair....I and others might listen then ;)

So far you have chose not to reply about this situation and that tells all.
ID: 4277 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John McLeod VII
Avatar

Send message
Joined: 27 Aug 07
Posts: 85
Credit: 405,705
RAC: 0
Message 4278 - Posted: 21 Jul 2008, 3:09:01 UTC - in response to Message 4277.  

Please keep personalities out of it.




Only if you stop threatening projects with xml stat manipulation or non-inclusion :)

I am aware (as well as you should be) that this is being done at the highest levels of BOINC to various project administrations.

Come back and talk cross project parity when that situation with Seti is fair....I and others might listen then ;)

So far you have chose not to reply about this situation and that tells all.

What I got was asses calling people that talk about cross party parity whiners.

If I were a moderator, much of this thread would go as flames.


BOINC WIKI
ID: 4278 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Jayargh
Avatar

Send message
Joined: 8 Oct 07
Posts: 289
Credit: 3,690,838
RAC: 0
Message 4279 - Posted: 21 Jul 2008, 3:10:51 UTC

I agree for once John....the original post needs to be deleted as non relevant.
ID: 4279 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John McLeod VII
Avatar

Send message
Joined: 27 Aug 07
Posts: 85
Credit: 405,705
RAC: 0
Message 4280 - Posted: 21 Jul 2008, 3:12:58 UTC

The stock applications at S@H are currently much better optimized than they were, and even the best optimized applications do not grant at 2 to one, but if I recall, only around 1.3 to one at the moment. Trust me on this. S@H and BOINC are being re-worked to make it much easier for projects to have optimized applicatons delivered to clients.

That still leaves some projects that are way out of line in granting credits.


BOINC WIKI
ID: 4280 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Jayargh
Avatar

Send message
Joined: 8 Oct 07
Posts: 289
Credit: 3,690,838
RAC: 0
Message 4281 - Posted: 21 Jul 2008, 3:16:18 UTC - in response to Message 4278.  



What I got was asses calling people that talk about cross party parity whiners.

If I were a moderator, much of this thread would go as flames.



Yes John lets not discuss the facts ...just call me an ass or a whiner
ID: 4281 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Jayargh
Avatar

Send message
Joined: 8 Oct 07
Posts: 289
Credit: 3,690,838
RAC: 0
Message 4282 - Posted: 21 Jul 2008, 3:19:15 UTC - in response to Message 4280.  
Last modified: 21 Jul 2008, 3:26:30 UTC

The stock applications at S@H are currently much better optimized than they were, and even the best optimized applications do not grant at 2 to one, but if I recall, only around 1.3 to one at the moment. Trust me on this. S@H and BOINC are being re-worked to make it much easier for projects to have optimized applicatons delivered to clients.

That still leaves some projects that are way out of line in granting credits.


Why stick to the 3rd party app? What is the motive? A score of projects do it internally and everyone benefits...its not just easier...everyone benefits....why does Seti promote a special club?

So again with an unfair system 1.3 to 1 will get most real crunchers to the highest credit which is SETI if you were to get your way in your proposal to all projects.

Why do I not see this crusade against LHC one of the lowest and barely mentioned ,but mentioned by BOINC admin as being low? Are you prejudiced attacking the high and leaving the low alone?
ID: 4282 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John McLeod VII
Avatar

Send message
Joined: 27 Aug 07
Posts: 85
Credit: 405,705
RAC: 0
Message 4283 - Posted: 21 Jul 2008, 3:26:10 UTC - in response to Message 4274.  

Having approximately the same credits per hour for a particular host across projects is what is SUPPOSED to happen in BOINC. If people are tired of hearing about this, then please ask the project administrators to get the calculation right rather than being proud of running off people that keep making that claim.

Projects that generate grossly too high credit claims should be removed from the cross project stats sites as this is just a way of stupidly inflating your cross project credit scores.

The long tasks really should be granting about 50 CS per task. Granting more is anti-social of the project administrators. The project administrators should look at what other projects are granting per hour for a particular host and attempt to match that here.



Participants who continually badger projects on this subject aka "Credit Police" should be deleted from the cross project stats IMHO.

Any stats sites that attempt to modify xml stats will be boycotted and thus loose their user base and advertising revenue thus nullifying this superior Boinc/Seti attitude that they are in contol...."Its the users stupid" to paraphrase which have control sorry ;) I will 'volunteer to lead the boycott'...now banish me.

Its nice to know that someone thinks they can regulate 50,80,100 projects this way ;)

If I am not mistaken doesn't the Milkyway Project use optimized apps? Thereby conforming to the side rule governing such apps? Thereby conforming to the prescribed Gestapo credit for the original un-optimized app?

Seti opened the credit Pandora's box with their allowed use and abuse of optimized applications....and now they attempt this strong arm approach again to squash it because it detracts from their user base?

Oh gosh darn I just realized Seti just changed their credit awards and philosophy ...guess this means all the good little projeckts must fall in line to das furor....or else!


Yes you do detect bitterness on my behalf based on years of cross project parity shenanigans....,political at best .....controlling at worst.

Yes JMVII they should evaluate against Cosmology as Milkyway is way too low in comparasin, and Milkyway has said optimized apps exonerating it.Please post this over there as I am ready fer ya ;)

They send one minion after another don't they? At least this one didn't try to hijack a thread.

Hmm. Boycotted by a project with merely 5K members and only 3K active? Hardly much of a threat. They might make that up by people feeling that it was more fair to stop reporting projects that were over granting.

And to the best of my knowledge, S@H changed the credit grants well over a year ago - when more optimization was added to the stock client. The grants were shrunk so as to keep approximately the same number of credits per hour on a similar host. This hardly feels like a change of philosophy to me.

The cross project parity discussion was started long before you joined this project, and probably long before you joined ANY BOINC project.

The principal designer of BOINC has come up with an idea that would ENFORCE cross project parity, but would also cause long term DEFLATION in credit grants. I would really hate for him to get tired of the cross project credit parity debates and actually implement his idea. The problem is that a CPU that is granted 100 CS / day NOW would be granted about 50 CS / day 18 months from now when CPU speeds double - this is not my idea of how to fix the problem, but I believe that the cross project parity problem will be fixed somehow. I would much rather have a political solution that involves telling projects that are way out of line with credit grants than a draconian measure that keeps reducing credit grants into the future.


BOINC WIKI
ID: 4283 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John McLeod VII
Avatar

Send message
Joined: 27 Aug 07
Posts: 85
Credit: 405,705
RAC: 0
Message 4284 - Posted: 21 Jul 2008, 3:31:18 UTC - in response to Message 4281.  



What I got was asses calling people that talk about cross party parity whiners.

If I were a moderator, much of this thread would go as flames.



Yes John lets not discuss the facts ...just call me an ass or a whiner

I was not calling you a whiner. Look at the second post in the thread - the one by voltron. That post in particular is NOT discussing the facts.


BOINC WIKI
ID: 4284 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John McLeod VII
Avatar

Send message
Joined: 27 Aug 07
Posts: 85
Credit: 405,705
RAC: 0
Message 4285 - Posted: 21 Jul 2008, 3:32:18 UTC - in response to Message 4279.  

I agree for once John....the original post needs to be deleted as non relevant.

It is really hard for the first thread to be irrelevant to the topic of the thread.


BOINC WIKI
ID: 4285 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Jayargh
Avatar

Send message
Joined: 8 Oct 07
Posts: 289
Credit: 3,690,838
RAC: 0
Message 4286 - Posted: 21 Jul 2008, 3:35:25 UTC - in response to Message 4283.  
Last modified: 21 Jul 2008, 4:00:24 UTC


Hmm. Boycotted by a project with merely 5K members and only 3K active? Hardly much of a threat.


Must have you flustered John as I would expect a better response than that!

Are all participants of Cosmology only connected to 1 project? Do you think team members would not support unfair treatment? Force away......you will never achieve anything near submission going into project boards arguing this same point....The rally cry may sound at such dismissive behaviour! You think the usership here will listen with those kind of inflammatory comments?

We and Cosmology have now been dismissed by Boinc as hardly much of a threat....ok I am done ...I refuse to respond to this type of Geastapoism.

You never answered as to why SETI needs to have a private club of optimized...the crux of my arguement?

I contend DA wants everyone else the same so Seti can still have an unfair advantage over every other project for credit no matter how miniscule that advantage turns out to be.

I will never buy into cross project parity until Seti cleans its optimized act up ...I am not the only user of this persuasion ya know.
ID: 4286 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John McLeod VII
Avatar

Send message
Joined: 27 Aug 07
Posts: 85
Credit: 405,705
RAC: 0
Message 4287 - Posted: 21 Jul 2008, 3:39:58 UTC - in response to Message 4286.  
Last modified: 21 Jul 2008, 3:44:07 UTC


Hmm. Boycotted by a project with merely 5K members and only 3K active? Hardly much of a threat.


Must have you flustered John as I would expect a better response than that!

Are all participants of Cosmology only connected to 1 project? Do you think team members would not support unfair treatment? Force away......you will never achieve anything near submission going into project boards arguing this same point....The rally cry may sound at such dismissive behaviour! You think the usership here will listen with those kind of inflammatory comments?

We and Cosmology have now been dismissed by Boinc as hardly much of a threat....ok I am done ...I refuse to respond to this type of Geastapoism.

I call Godwin's Law.

Did you read the threat about a change to the server code to enforce cross project credit parity and what that would do to credits long term? I hope we have talked him out of the idea, but Dr. A. feels the need for cross project parity very strongly, and he is the one that controls the BOINC code. Like I stated earlier, I would prefer that all projects grant credit equitably rather than have that rather draconian solution forced upon us.


BOINC WIKI
ID: 4287 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Bill & Patsy
Avatar

Send message
Joined: 7 Jul 08
Posts: 47
Credit: 13,629,944
RAC: 0
Message 4288 - Posted: 21 Jul 2008, 3:52:19 UTC

Jeff is right, John, and you know it. That focus on political correctness (PC) rather than science is why I (and others) left SETI a year ago. (My SETI share is now below 0.03%). This was at the time significant optimizations were incorporated into the standard apps, as you and Jeff mentioned. This resulted in more science being done - more results. At work, I get paid for the work I do (the results I produce), not my effort. But your PC bias rejected what the rest of us were saying, with SETI insisting that the PC thing was to reward effort rather than scientific contribution. So SETI degraded the reward scale and we left, because BOINC is a free enterprise community.

I'm chagrined to see, a year later, that you are still peddling the propaganda that effort is all that matters. It's just plain irrational to maintain that two identical machines producing unequal outputs should nevertheless be looked at as doing the same work! Give it up, John!! The Free World doesn't pay much for effort!

There are two principle criteria I look for in a project. First and foremost is its scientific worth (based subjectively, of course, on the sciences I value most). Secondly is how much the project wants my participation. SETI surely wins on the first, but a year ago failed miserably on the second.

Lastly, John, concerning "personalities", I would remind you that there are two ways to have the tallest building in town. One is to tear the others down. Please stop going around trying to tear down the competition. You wouldn't have this "problem" if you hadn't started this mess in the first place by manipulating SETI's own recognition scale to reward effort rather than results. Fix the moat in your own eye first.

Oh yeah. And it's not a good idea to threaten other BOINC projects. BOINC is not the only distributed computing paradigm around, and if your "principal designer" becomes an enforcer, you could see BOINC unravel real fast.
--Bill

ID: 4288 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John McLeod VII
Avatar

Send message
Joined: 27 Aug 07
Posts: 85
Credit: 405,705
RAC: 0
Message 4289 - Posted: 21 Jul 2008, 4:02:48 UTC - in response to Message 4286.  


Hmm. Boycotted by a project with merely 5K members and only 3K active? Hardly much of a threat.


Must have you flustered John as I would expect a better response than that!

Are all participants of Cosmology only connected to 1 project? Do you think team members would not support unfair treatment? Force away......you will never achieve anything near submission going into project boards arguing this same point....The rally cry may sound at such dismissive behaviour! You think the usership here will listen with those kind of inflammatory comments?

We and Cosmology have now been dismissed by Boinc as hardly much of a threat....ok I am done ...I refuse to respond to this type of Geastapoism.

You never answered as to why SETI needs to have a private club of optimized...the crux of my arguement?

BOINC is not well designed to distribute optimized applications to different clients of the same basic CPU and OS. It was made barely possible when the CPU optimizations were reported back. Work is in progress for the BOINC client and server to make this relatively easy, and S@H is, I believe waiting for that. The anonymous client was implemented long before the detection code was implemented in the client to distinguish which FPU extensions were implemented. The anonymous client was implemented so that people could develop applications for hardware that the project could not devote the manpower to implement the application for (i.e. OS2). It has been used for this. Some users also figured out that the same mechanism could be used to implement optimized applications. S@H has implemented some of the optimizations that have been developed into the stock client. Others have not been incorporated. There was an enormous debate about credits granted to optimized apps, and another when S@H incorporated some optimization and cut the credit grant to keep the credits / hour on hosts with stock applications about the same.

Basically what has happened has been that better hardware gets more credit per hour (the optimized applications count as "better hardware"). However, when the stock application gets optimized that does not count as "better hardware" for everyone and the credit grant was reduced to match the increased performance. I expect that when (may be later as Astropulse is absorbing much of the programming time) optimized apps per FPU extension start being used the credit request will be reduced again to reflect the reduced FLOPS count and to keep the average credits per hour about the same.

BTW, did you know that about 1% of the time the optimized applications at S@H fail to generate the correct result and therefore get no credit granted? This is one reason that not all of the optimizations were incorporated into the stock client. Another is, of course, that the BOINC client was not reporting the FPU extensions and therefore the server could not send a specifically optimized client.


BOINC WIKI
ID: 4289 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John McLeod VII
Avatar

Send message
Joined: 27 Aug 07
Posts: 85
Credit: 405,705
RAC: 0
Message 4290 - Posted: 21 Jul 2008, 4:10:40 UTC - in response to Message 4288.  

Jeff is right, John, and you know it. That focus on political correctness (PC) rather than science is why I (and others) left SETI a year ago. (My SETI share is now below 0.03%). This was at the time significant optimizations were incorporated into the standard apps, as you and Jeff mentioned. This resulted in more science being done - more results. At work, I get paid for the work I do (the results I produce), not my effort. But your PC bias rejected what the rest of us were saying, with SETI insisting that the PC thing was to reward effort rather than scientific contribution. So SETI degraded the reward scale and we left, because BOINC is a free enterprise community.

I'm chagrined to see, a year later, that you are still peddling the propaganda that effort is all that matters. It's just plain irrational to maintain that two identical machines producing unequal outputs should nevertheless be looked at as doing the same work! Give it up, John!! The Free World doesn't pay much for effort!

There are two principle criteria I look for in a project. First and foremost is its scientific worth (based subjectively, of course, on the sciences I value most). Secondly is how much the project wants my participation. SETI surely wins on the first, but a year ago failed miserably on the second.

Lastly, John, concerning "personalities", I would remind you that there are two ways to have the tallest building in town. One is to tear the others down. Please stop going around trying to tear down the competition. You wouldn't have this "problem" if you hadn't started this mess in the first place by manipulating SETI's own recognition scale to reward effort rather than results. Fix the moat in your own eye first.

Oh yeah. And it's not a good idea to threaten other BOINC projects. BOINC is not the only distributed computing paradigm around, and if your "principal designer" becomes an enforcer, you could see BOINC unravel real fast.

Sorry, I do NOT believe that either you or Jeff is correct, and I certainly do not know that you are correct. If that were the case, there would not be a discussion.

Certainly, you can pick based on scientific value. What you are saying is that you are selecting based on how many credits per hour you are granted (how much the project values you). Having projects grant whatever they feel like is a way to have an inflationary race:

Project 1 needs more participants so it raises the credits granted.
Now project 2 does the same.
Now back to project 1...

I do not see this as a good scenario, but apparently you do.




BOINC WIKI
ID: 4290 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Jayargh
Avatar

Send message
Joined: 8 Oct 07
Posts: 289
Credit: 3,690,838
RAC: 0
Message 4291 - Posted: 21 Jul 2008, 4:30:19 UTC - in response to Message 4289.  
Last modified: 21 Jul 2008, 5:29:14 UTC


Hmm. Boycotted by a project with merely 5K members and only 3K active? Hardly much of a threat.


Must have you flustered John as I would expect a better response than that!

Are all participants of Cosmology only connected to 1 project? Do you think team members would not support unfair treatment? Force away......you will never achieve anything near submission going into project boards arguing this same point....The rally cry may sound at such dismissive behaviour! You think the usership here will listen with those kind of inflammatory comments?

We and Cosmology have now been dismissed by Boinc as hardly much of a threat....ok I am done ...I refuse to respond to this type of Geastapoism.

You never answered as to why SETI needs to have a private club of optimized...the crux of my arguement?

BOINC is not well designed to distribute optimized applications to different clients of the same basic CPU and OS. It was made barely possible when the CPU optimizations were reported back. Work is in progress for the BOINC client and server to make this relatively easy, and S@H is, I believe waiting for that. The anonymous client was implemented long before the detection code was implemented in the client to distinguish which FPU extensions were implemented. The anonymous client was implemented so that people could develop applications for hardware that the project could not devote the manpower to implement the application for (i.e. OS2). It has been used for this. Some users also figured out that the same mechanism could be used to implement optimized applications. S@H has implemented some of the optimizations that have been developed into the stock client. Others have not been incorporated. There was an enormous debate about credits granted to optimized apps, and another when S@H incorporated some optimization and cut the credit grant to keep the credits / hour on hosts with stock applications about the same.

Basically what has happened has been that better hardware gets more credit per hour (the optimized applications count as "better hardware"). However, when the stock application gets optimized that does not count as "better hardware" for everyone and the credit grant was reduced to match the increased performance. I expect that when (may be later as Astropulse is absorbing much of the programming time) optimized apps per FPU extension start being used the credit request will be reduced again to reflect the reduced FLOPS count and to keep the average credits per hour about the same.

BTW, did you know that about 1% of the time the optimized applications at S@H fail to generate the correct result and therefore get no credit granted? This is one reason that not all of the optimizations were incorporated into the stock client. Another is, of course, that the BOINC client was not reporting the FPU extensions and therefore the server could not send a specifically optimized client.


John thank you for that answer but my question still has not been answered in that for example Einstein had a 3rd party app come about the same reasons as you describe as the project did not have the ability to do so....but they incorporated it into an app everyone receives....why can't the 3rd party be eliminated ,have the project incorporate all it sees fit into what is sent out as an app and credit those individuals who contributed to the code ?

Politics and control questions have not been answered...why the continued private club as I call it?

Another question not answered...Why no postings (et al harrasement) of the lowest undergranting projects like LHC? Why go after the 'ridiculous high' and not the low if not political?

Another scenario...a project has 1gb memory requirements....tad high for most hosts....should not that project be allowed to compensate at a higher credit level due to the ensuing computation errors at unexpected times? How much is enough....how much is too much? This is why it won't work, projects with super long or tough tasks to complete will never have a chance!

I see no allowance for this in the 'enforcement agenda'

I predict another platform will come along to compete with Boinc if all this comes to pass.
ID: 4291 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
1 · 2 · 3 · 4 . . . 9 · Next

Message boards : Number crunching : Credit Calculations.

©2024 Astroinformatics Group