(...) Given we have scientists running the project, I would ask where is the offline analysis which supports the proposition that there is nothing wrong with the credit system ???

Sorry, i don't have an answer, but i like this question very much! %))

---

Aloha, Uli

(...) Given we have scientists running the project, I would ask where is the offline analysis which supports the proposition that there is nothing wrong with the credit system ???

Sorry, i don't have an answer, but i like this question very much! %))

Sorry, you need to reformulate the question. Note that you normally cannot 'prove' a negative!


:-p

Happy fast crunchin',
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

Hey - has anybody else noticed that the credit for Seti 7 WUs has dropped significantly the last day or two? I am seeing lots of WUs giving 30-40 credits, and far fewer in the 80-100 range, where it seemed to have stabilized for a while before.

I don't think it is a shorty storm, either, as they don't seem to be running less time (at least as far as I can tell).

Thanks for any info.

---

(...) Given we have scientists running the project, I would ask where is the offline analysis which supports the proposition that there is nothing wrong with the credit system ???

Sorry, i don't have an answer, but i like this question very much! %))

Sorry, you need to reformulate the question. Note that you normally cannot 'prove' a negative!

Really? When I moved to the UK the Swiss asked me to prove I'd never been married before they'd pay out my pension/superannuation pot.

---

Not sure anyone knows this for sure, but many are saying it has normalised.

In looking at tasks going through the i7, enhanced wus are still receiving significantly more credit than v7 wus (even though enhanced wus take ~50% less time to crunch). Enhanced and AP seem to be roughly equivalent whilst v7 stands out like the proverbial sore thumb. Simple logic would dictate that if a wu takes twice as long to crunch then one should expect to see twice the credit. I am not seeing this with v7 wus compared to enhanced. Given we have scientists running the project, I would ask where is the offline analysis which supports the proposition that there is nothing wrong with the credit system ???

And you would be wrong. "Standard" WU reported in the 90-110 range on average with V6. Shorties were in the 30s.

Now right after the switch over those values plummeted and it just took a while for the new averages, since the standard apps are now more optimized than before.

But now, looking at my valid WUs show that once again we are back up to those pre-V7 values.

That said looking at your enhanced show some 130, 160 even a 180 credit WU being done. They have rather uncommon angle range in the 0.31-0.35 range. Glancing over your V7 results show that V6 and V7 units with the more common angle range of 0.41 are both crediting in the 100-105 range.

---

"Life is just nature's way of keeping meat fresh." - The Doctor

Glancing over your V7 results show that V6 and V7 units with the more common angle range of 0.41 are both crediting in the 100-105 range.

This is where theory & reality diverge.
v7 work takes longer to process than v6- so even though the amount of credit per WU is the same, more work is being done so in actual fact the amount of credit for work done has fallen (as RAC shows).
If the amount of credit paid was based on work done (as it was intended to) then the amount of credit paid for v7 work should be higher to offset the longer processing time. The result of that being the credit per hour would have remained unchanged, so RAC would have stayed (roughly) the same.
It didn't, because Credit New pays on theoretical processing efficiency, not for the work that is actually done.

---

Grant
Darwin NT

Interesting observation. I just put another GTX 660 in one of my crunchers taking it up to 3 GPU. I noticed that about 5 minites of running my RAC dropped from 87,000 to 86,000. Coincedence??

Interesting observation. I just put another GTX 660 in one of my crunchers taking it up to 3 GPU. I noticed that about 5 minites of running my RAC dropped from 87,000 to 86,000. Coincedence??

Probably a coincidence, since obviously 3 GPUs should produce more than 2 ;) Kindof tongue in cheek, did you take ~1000 seconds (~16mins40secs) to power down, install, then power up & resume crunching (with 3 cards) ?

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

I took slightly longer as I had to hang my power supply out side my box. Looks terrible but it works

This is where theory & reality diverge.
v7 work takes longer to process than v6- so even though the amount of credit per WU is the same, more work is being done so in actual fact the amount of credit for work done has fallen (as RAC shows).
If the amount of credit paid was based on work done (as it was intended to) then the amount of credit paid for v7 work should be higher to offset the longer processing time. The result of that being the credit per hour would have remained unchanged, so RAC would have stayed (roughly) the same.
It didn't, because Credit New pays on theoretical processing efficiency, not for the work that is actually done.

It is my understanding that the CreditNew is based on the "standard" CPU app. Now I believe that the V7 standard app is now based on the optimized app while the V6 standard app wasn't. Now what if the average time of optimized processing plus auto-correlation of the current V7 standard CPU app takes about the same average time of the unoptimized standard CPU app under V6? Wouldn't that mean the credits rewarded be similar? And since it's the standard CPU app that determines the baseline credit rate everything else is now based on it.

Cursory examination of the few standard seti_enhanced CPU app results seem to indicate this. Less than a 5% loss in RAC. So to those who never used optimized apps have seen little change in their RAC while those who do see a major decrease.

Of course those of us who were running optimized apps under V6 see WUs take longer, GPU especially, but that doesn't matter since it's the standard app that sets, well, the standard. It's just now our optimized CPU app isn't that much better than the standard CPU app and while our optimized GPU app suffers by a lot, it's still many times faster than a CPU app.

We've gotten such a sense of credit rate elitism after years of being significantly faster than the standard app that when the standard app catches up to us we get all boo-hooey over our perceived loss of superiority. We are just angry that we are no longer so much better than someone running a standard app. Well we have to learn to get over it.

Edit: Also I believe that the auto-correlation code does not add to the reported flop count. So even though the app is doing more, it's not reported as doing more.

---

"Life is just nature's way of keeping meat fresh." - The Doctor

My global ranking by RAC is pretty similar to what it was before the switchover, about 14,000.

http://boincstats.com/en/charts/-1/user/racRankDay/2522086/chart.png

The only machine that I saw a drop in was my I7 920 with a GTS 250. Before the switch It was running between 10 and 12K rac. That was with lunatics op apps.

After the switch I dropped to 3k Rac. It is now close to 6K. Will it get back to 10K? Who knows and I dont care.

However before the switch my two I7 3770 each with a 550 Ti running standard apps were around 7K rac. After the switch I installed lunatics op apps and each rig is now closing in on 16k Rac. And that is running HT with all 8 cores running. ( I will free a core when the Rac gets stable just to see if it makes a differance.)

Id like to pop a 550 Ti in the 920 to see what would happen but the card wont fit in the case.

In my case the Rac has increased on two machines and the other one is going up for who knows how long. And I have never cherry picked. I crunch what i get.

---

[/quote]

Old James

Cursory examination of the few standard seti_enhanced CPU app results seem to indicate this. Less than a 5% loss in RAC. So to those who never used optimized apps have seen little change in their RAC while those who do see a major decrease.

Of course those of us who were running optimized apps under V6 see WUs take longer, GPU especially, but that doesn't matter since it's the standard app that sets, well, the standard. It's just now our optimized CPU app isn't that much better than the standard CPU app and while our optimized GPU app suffers by a lot, it's still many times faster than a CPU app.

Which shows that the system is broken.
The whole point of Credits was to recognise work done.
Since the stock application is now using optimisations it should result in the stock application now earning similar credit to what the previous optimised application did. It shouldn't result in less credit being granted for the same work being done, it should have resulted in more credit being granted as the stock application is now doing more work.
The stock application is now doing more work- it should provide more credit to recognise that. For those that were previously using the optimised application the amount of credit received should have remained about the same (the newly introduced auto correlations would probably have resulted in some drop in credit- not the 40% or more that has occured). Credit deflation for those that used the optimised application should not have been the result.

---

Grant
Darwin NT

+1

It is not credit elitism as Keith suggested to be the reason for the grunts.
It is that instead of increasing the credits awarded to the people who are now using stock, but optimised, applications, the credit for the people who were on optimised has fallen. And it should have stayed the same (roughly).
If only the "normal peasant" people had increased their RAC the aristocratic ones would not have had any reason to complain. And they would not have.

Happy crunching!

Sleepy
________________________________________________________
100 g of ham are 100 g of ham and should be paid as such

---

Which shows that the system is broken.
The whole point of Credits was to recognise work done.

The only ones who not see that is Dr. A. & Co.
Did you ever tried to make a blind person see?

In the V6 era my RAC was ~450k now with v7 ~235k, who could say it not falls?
and all my host allready use x41zc (cuda50 & 42) on the v6 era as they do now.

They actualy make the same works as before (maybe even a little more due correlation), but some blind people decide they will receive a lot less credit per compleated WU than before or at least makes nothing to change that aberration called creditnew!

---

Keith White wrote:

It is my understanding that the CreditNew is based on the "standard" CPU app. Now I believe that the V7 standard app is now based on the optimized app while the V6 standard app wasn't. Now what if the average time of optimized processing plus auto-correlation of the current V7 standard CPU app takes about the same average time of the unoptimized standard CPU app under V6? Wouldn't that mean the credits rewarded be similar? And since it's the standard CPU app that determines the baseline credit rate everything else is now based on it.

Now that would be interesting to know, is the v7 stock app actually as fast per WU as v6? Because than it would be correct, that we get about the same credit per WU as with v6 or even less if the v7 stock app is faster per WU than stock v6. Has somebody tested that for example by running both apps on the same WU?



Grant (SSSF) wrote:

Since the stock application is now using optimisations it should result in the stock application now earning similar credit to what the previous optimised application did. It shouldn't result in less credit being granted for the same work being done, it should have resulted in more credit being granted as the stock application is now doing more work.
The stock application is now doing more work- it should provide more credit to recognise that. For those that were previously using the optimised application the amount of credit received should have remained about the same (the newly introduced auto correlations would probably have resulted in some drop in credit- not the 40% or more that has occured). Credit deflation for those that used the optimised application should not have been the result.

No. Credit should reflect the amount of computing power that a user donates to a project, not how much science the project is doing with it. If SETI would not be accepting optimized applications and one day they would release better optimized applications by themselves, we would also not get more credit, we would get less credit per WU but we would do more WUs per day which in case of properly working credit system should result in same RAC, even if more science was done.

---

No. Credit should reflect the amount of computing power that a user donates to a project, not how much science the project is doing with it. If SETI would not be accepting optimized applications and one day they would release better optimized applications by themselves, we would also not get more credit, we would get less credit per WU but we would do more WUs per day which in case of properly working credit system should result in same RAC, even if more science was done.

Which completely goes against the whole idea of granting credit.
With the opriginal project, the number of WUs processed was the only way of keeping track of the work done. However significant amounts of what was returned were corrupt, invalid, or the same very short WU crunched over & over again.
To stop cheating & better reflect the work done, the Credit system was devised.

If it was to accurately show the computing power devoted to the project, that would be reflected in the amount of work done. Which is what Credit is meant to do, but no longer does.
If you have more powerfull hardware, or a more effcicent application, you should receive more credit per hour. A more optimised stock application should not result in a drop in credit for those that have always been using such applications, but more credit for the stock application- reflecting the greater amount of work done.
Inflation isn't good, niether is deflation.

---

Grant
Darwin NT

The only ones who not see that is Dr. A. & Co.
Did you ever tried to make a blind person see?

Has any of the complainers contacted Dr. A. to try and
get him to 'see' things their way?

I'd bet money he doesn't read this forum or thread,
in which case nobody is trying to make the 'blind' man
see anything, they're just kvetching to each other, which
will accomplish exactly nothing.

The only ones who not see that is Dr. A. & Co.
Did you ever tried to make a blind person see?

Has any of the complainers contacted Dr. A. to try and
get him to 'see' things their way?

I'd bet money he doesn't read this forum or thread,
in which case nobody is trying to make the 'blind' man
see anything, they're just kvetching to each other, which
will accomplish exactly nothing.

The are actualy few e-mails sended by me and some others about that, but not see anyone allready answered.

---

Good. I would think that enough people showing
there's a problem would at least get a reply of some sort.

Of course, it might not be a reply that anyone likes,
but at least we'd know what's being done.