Page 1 of 1

Cheater suspected!

Posted: Tue Jan 06, 2009 9:35 pm
by dehnhaide
Hi everyone!

I am writing here in order to draw your attention towards some situation that I find at least weird, or even more, I suspect cheating. I am talking here about user "Olorin" from my team "Folding@Romania" (team no. 12713).
Things are like this: this gus has not been folding at all for the past year or so and now he has popped up with some very curious PPDs for his WUs.

Code: Select all

http://folding.extremeoverclocking.com/user_summary.php?s=&u=77023
Looking at this stats you will notice that he's got 14 times (15,360PPDs for 8WUs and 11,520PPDs for 6WUs in a row) the same huge & magic number 1920 PPDs per 1 WU!
I have been folding for more than a year (using more than 10 different processors, platforms & architectures) and I have never been able to get such an huge number of PPDs for a WU and more so, make it consequently. I have also looked to other teams and team member stats and did not notice such spectacular "round" numbers. I could see that people, even when folding with 1 or 2 CPUs will get almost each time a different amount of PPDs as the client itself shifts the project number almost every new run! Am I wrong?

Last of his "prize" goes for: 01.06, 12pm --> 25,344 PPDs --> for 14 WU making it, again, a stunning average of 1810.28 PPD per WU.
People, I don't want to sound paranoid or dumb but just PLEASE clear out this one for me! I just want to find out how can I set my "whatever folding client" with "whatever options" to run projects that would award such a huge number of PPDs. I would like to find out if there's any miracle CPU capable of crunching WUs of such PPD value - I would buy one or more for myself!

Don't get me wrong: I have started folding for one sole reason and that was helping science. I still enjoy every single PPD that I get for my work and believe me there is effort I have put on gathering 14 CPUs to run folding@home on for now. I just want to make sure that what my team "folding@romania" does is pure folding effort and NO cheating. Can somebody please check deeper on this situation.
I am also available to apologize if my case is wrong!

Thank you for your time.

Re: Cheater suspected!

Posted: Tue Jan 06, 2009 9:40 pm
by bollix47
There are a lot of WUs that are worth 1920 points when using SMP. ;)

Re: Cheater suspected!

Posted: Tue Jan 06, 2009 9:45 pm
by ChelseaOilman
bollix47 wrote:There are a lot of WUs that are worth 1920 points when using SMP. ;)
Correct. He's running the SMP client under Linux and getting the A2 WUs.

Re: Cheater suspected!

Posted: Tue Jan 06, 2009 9:55 pm
by ChasR
An Intel quad running native Linux produces about 1800 points per day per GHz (Q6600, newer quads slightly higher) on A2 SMP WUs. So running stock, a single Q6600 will complete 2.25 1920 point WU/day. No cheating going on.

Re: Cheater suspected!

Posted: Tue Jan 06, 2009 10:03 pm
by MoneyGuyBK
ChelseaOilman wrote:
bollix47 wrote:There are a lot of WUs that are worth 1920 points when using SMP. ;)
Correct. He's running the SMP client under Linux and getting the A2 WUs.
@ dehnhaide
Olorin may have added/utilized a server of sorts... or someone's super-pc...
You should be glad that Olorin is contributing those hefty points to your team, if you ever manage to fire him/her, I am certain many teams will love to add that production to theirs ;) :P








Peace

Re: Cheater suspected!

Posted: Tue Jan 06, 2009 10:23 pm
by PlayLoud
ChasR wrote:An Intel quad running native Linux produces about 1800 points per day per GHz (Q6600, newer quads slightly higher) on A2 SMP WUs. So running stock, a single Q6600 will complete 2.25 1920 point WU/day. No cheating going on.
Yup, that's right on. My OC'd Q6600 (3.2) finishes 3x A2 WUs (1920 points each) per day. I get about 5800ppd.

Sometimes I get a REALLY big WU, at 3840 points. Instead of 8 hours, it takes 16 hours to complete (same PPD obviously).

Re: Cheater suspected!

Posted: Tue Jan 06, 2009 11:21 pm
by dehnhaide
Guys,

Thank you very much for your prompt and professional answers.
I don't feel ashamed at all for questioning Olorin's virtue on folding. I just feel happy and relieved now that I know he's doing it good, in huge amounts and, most of all, fair.
I am more than pleased to hear that the linux client could fold so good on SMP. Gotta change to 64bits OS pretty soon.

Kind regards.

Re: Cheater suspected!

Posted: Wed Jan 07, 2009 12:17 am
by Tobit
Can't wait to get my new Core i7 Linux workstation online. :)

Re: Cheater suspected!

Posted: Wed Jan 07, 2009 12:40 am
by divery4eyes
I shall hasten no more. This thread has convinced me into getting the quad. :ewink:

Re: Cheater suspected!

Posted: Sun Apr 19, 2009 10:45 am
by Olorin
Dear dehnhaide,

I was searching the net for stats on my team and I stumbled over this thread by accident. I am "the" Olorin contribuding to the folding@romania team, and indeed, as the others have mentioned (and convinced you) I am using the linux SMP folding engine. As for the machines I am using: they are the unused blades of the cluster I am running at my University. I am "only" using 16 dual-proc quad-core Intel(R) Xeon(R) CPU E5405 @ 2.00GHz. We are running RedHat 4.7. I can only say the linux SMP is worth it.

Regards,

Olorin