Prime95/SP2004 cheating

[Kunaak]

you might want to consider changing your rules alittle.
see prime 95 is one of the easiest things you can possibly cheat on.

I would suggest showing Cheat Free 32M shots atleast, with the Prime95 shots.

theres always someone that wants the recognition for something they never did.

heres a screen shot I just did.
took me about 10 seconds to do this.
obviously my prime95 was just opend, and theres no way possible to have 144 hours of stability when I just opend the program, but its easy to fool someone with it.
just let it run for a half hour or so, then fake the rest.

(don't ask how to cheat on Prime, cause I don't give out info, also, if you know how you can do it, don't go around telling people how to do it)

you need something more then prime 95 to prove stability, cause it's just so dang easy to cheat on.

I would suggest the Cheat Free version Mod 1.4 or newer, 32M, and 3dmark2001SE Links.
Those 2 things should help bring alot of credit to a claimed 24 stable shot.

http://www.xtremesystems.org/forums/attachment.php?attachmentid=30605&stc=1

[agenda2005]

Kunak,
Can you cheat the Johnny Lee stress prime? Try and let us see.

[Kunaak]

give me a link to it.

I am pretty sure I can.

EDIT:

Used Google to find it.

http://sp2004.fre3.com/download.htm

I like this program alittle better then Prime95, when used for actual stability testing, but it's just as easy to cheat on. takes almost no effort.

http://www.xtremesystems.org/forums/attachment.php?attachmentid=30608&stc=1

[agenda2005]

Here you go,
http://sp2004.fre3.com/

[Faceless Rebel]

I dunno, I asked in OP that people be honest and not cheat. I'll take your suggestion into consideration, however, since Super PI 32M only takes <1/2hr so it's not a huge deal these days. Does anybody else think this is a good idea?

[agenda2005]

Alright, I was trying to see that possibility myself, but I didn't want to stop my system from running prime. I may have to report this to the Aurthor to fix the problem.



Can you cheat on the CPU speed reported as well?

[conrad.maranan]

Sorry if this thread is not heading off in the right direction. But I, too, would like to add that anyone with basic Photoshop skills can easily edit Prime along with a number of other benchmarks. As stated time and time again, it is sad that it has come down to the point where people actually have to cheat their way into receiving praise from their peers. We need to follow the footsteps of our overclocking friends in Japan who test with integrity.

Perhaps a new version of Prime with a verification code similar to CPU-Z is in order. ;)

[Faceless Rebel]

Can you cheat on the CPU speed reported as well?

Well, he's making it report 4059mhz, so I'd say yes. :)


Sorry if this thread is not heading off in the right direction. But I, too, would like to add that anyone with basic Photoshop skills can easily edit Prime along with a number of other benchmarks. As stated time and time again, it is sad that it has come down to the point where people actually have to cheat their way into receiving praise from their peers. We need to follow the footsteps of our overclocking friends in Japan who test with integrity.

Perhaps a new version of Prime with a verification code similar to CPU-Z is in order. ;)

The Japanese do a lot of things with integrity. It's much more fun to play MMOGs with Japanese than Americans, for example. (If you're looking for an MMOG with lots of Japanese you can play with, Final Fantasy XI will interest you.) I remember when I played in the Bomberman Online free open beta, which was Japan only but you could download the client and a patch to make it work with English WinXP, the Japanese were all very nice and respectful. It's unfortunate that the American societal mentality is screw everyone as often as you can to get ahead, but there's nothing you or I can do about it. I can only ask politely that people do not cheat, and if they are suspected of cheating they either prove themselves or risk the banstick. I'd rather not have to look over everyone's shoulder to make sure they're not cheating... :)

[Kunaak]

cheating the CPU speed would be very easy as well.
very few programs have any cheat detection system, or anticheating.

if you know the programmer of some program, and want to get it cheat tested and worked with, then have the guy PM me, and I will help him with the testing and such.
I've helped develope a few cheat free programs so far, with varied success, but somethings, better then nothing.

[agenda2005]

I SEE! In fact, it's quite easy to cheat that program. My screen shot is 100% legit because I have nothing to gain deceiving myself about my system stability.

[Kunaak]

Well, he's making it report 4059mhz, so I'd say yes. :)

haha....

actually thats right.
my CPU is a 3.6 ES at the moment running 14x290 at 5:4
that part is the only thing in that shot that isn't cheated.

however, I can cheat that just as easily.

[agenda2005]

Try and cheat the CPU speed reported as well as the start time and let's see please because I couldn't figure it out.

[bachus_anonym]

SP2004 shows CPU Frequency as the one PC was booted at. So, you can boot @ 2800MHz, run SP2004 @ 2000MHz and still has it reporting 2800MHz.

Coiple of weeks ago I was also thinking of shooting an email to the author with some suggestions for cheat-free version.

[agenda2005]

SP2004 shows CPU Frequency as the one PC was booted at. So, you can boot @ 2800MHz, run SP2004 @ 2000MHz and still has it reporting 2800MHz.

Coiple of weeks ago I was also thinking of shooting an email to the author with some suggestions for cheat-free version.

I was not thinking about that. I think I will stop posting SS of this program until all these issues are resolved. I will keep using it for my own personal consumption anyway. It's a shame though.

[conrad.maranan]

SP2004 shows CPU Frequency as the one PC was booted at. So, you can boot @ 2800MHz, run SP2004 @ 2000MHz and still has it reporting 2800MHz.
Out of curiosity (and I never figured this one out), why do you think SP2004 reported my clocks as 7MHz (6MHz x 376.0 est.) in this screenshot (http://webpages.charter.net/conrad-digital/3500VX/10x263MHzROCKSOLID.png)? That was an old run of mine when the VX was first introduced and I can assure you that the Prime run was legit. My actual clocks were 10x263MHz.
:confused:

[Kunaak]

the method bachus told does indeed work, however, it has a interesting side effect also...

it's virtually immune to cheat speeding. ;)
I had RM Clock at 11 ghz, but CPUZ and SP2004 were both showing the correct speed.
this is expected from CPUZ, but not from SP2004.
I am pretty sure I know why that is though...

[bachus_anonym]

SP2004 has this cool feature of periodically reading sensors of MBM, like every 5 minutes... So I believe It could be quite easy for the author to do the same with CPu Freq Detection, that will be done every 5min and printed along with Iterations.

This SP2004 has a potential to be quite good cheat-free test of stability, In my opinion...

[agenda2005]

Out of curiosity (and I never figured this one out), why do you think SP2004 reported my clocks as 7MHz (6MHz x 376.0 est.) in this screenshot (http://webpages.charter.net/conrad-digital/3500VX/10x263MHzROCKSOLID.png)? That was an old run of mine when the VX was first introduced and I can assure you that the Prime run was legit. My actual clocks were 10x263MHz.
:confused:

That's awfull :stick:

[bachus_anonym]

Out of curiosity (and I never figured this one out), why do you think SP2004 reported my clocks as 7MHz (6MHz x 376.0 est.) in this screenshot (http://webpages.charter.net/conrad-digital/3500VX/10x263MHzROCKSOLID.png)? That was an old run of mine when the VX was first introduced and I can assure you that the Prime run was legit. My actual clocks were 10x263MHz.
:confused:
It reported your FSB as 263-256=7MHz.... Remember, 3dMark2001 does same thing... On Neo2 or NF2 mobos...

[Kunaak]

actually from the method your talking about, I think it would make it virtually useless as a cheat free program...

what I think SP2004 right now is doing, is getting its speed from Windows itself.
not from bios or anything else.
this is why it wouldn't be changable, even with clockgen and such.
thats good for verifying an OC, if that speed is the same speed you booted up at.
but useless if you use clockgen in windows then wanted to use SP2004 to help verify a OC.

constant sampling of the CPU clock, would just make it update nice and fast so it would be nice and easy to cheat with.

to get a cheat free clock reading, you have to read directly from the Bios, and have zero dependancy on windows.

[conrad.maranan]

It reported your FSB as 263-256=7MHz.... Remember, 3dMark2001 does same thing... On Neo2 or NF2 mobos...
Wait. I'm confused. Where did you derive the value of 256 from? :confused:

[agenda2005]

actually from the method your talking about, I think it would make it virtually useless as a cheat free program...

what I think SP2004 right now is doing, is getting its speed from Windows itself.
not from bios or anything else.
this is why it wouldn't be changable, even with clockgen and such.
thats good for verifying an OC, if that speed is the same speed you booted up at.
but useless if you use clockgen in windows then wanted to use SP2004 to help verify a OC.

constant sampling of the CPU clock, would just make it update nice and fast so it would be nice and easy to cheat with.

to get a cheat free clock reading, you have to read directly from the Bios, and have zero dependancy on windows.

I discovered that too. SP2004 doesn't update CPU speed from clockgen. But Conrad SS with 7MHZ FSB is scary if such error exist.

[bachus_anonym]

I do not really know where the value of 256 came from (perhaps 8 bits allocated for FSB closk in detection engine :shrug: of 3DMark2001) but If you run 3dMark2001 at 256MHz FSB it would report it as 256. But 257MHz would be detected as 1MHz and so on.

So it looks like SP2004 uses exactly same procedure to detect CPU/FSB clocks...

EDIT: Anyway, here goes Faceless Rebel's thread :lol: We just trashed it real good :D

[conrad.maranan]

No, I'm going to split it, consolidate our off-topic posts into a thread named Prime95/SP2004 cheating, and then move it to XtremeSystems Forums > Software > Benchmarking Programs. ;)

[Faceless Rebel]

Thanks for splitting this thread off, what I'll do is edit my OP to link to this thread so people will see it. :)

[conrad.maranan]

No sweat, Faceless Rebel. ;)

Michal - Sorry for not replying right away since I was leaving the office to go home. Anyway, are you saying that anything after 256MHz throws off the whole CPU/FSB detection? I'm not much of a 3DMark bencher so I wasn't aware of this tidbit of information. Thanks for sharing, though.

[Kunaak]

I have a solution for "Hard to Cheat On" stability testing.

Again, Wizzard, has a solution, and it can be found in his Systool Program.

I tested this program in a few different ways, and I had a hard time cheating on it, see, with it's constant updating the list as to what passed and what didn't, I couldn't create a cheat that looked legit, cause it wouldn't match the next results...
thats very very good, for spotting cheats real fast.

manual cheating, and speed cheating did nothing to this program.

so it's about 90% cheat free, as far as I can tell. :D

http://systool.techpowerup.com

http://www.xtremesystems.org/forums/attachment.php?attachmentid=30616&stc=1

[jikdoc]

Well, he's making it report 4059mhz, so I'd say yes. :)



The Japanese do a lot of things with integrity. It's much more fun to play MMOGs with Japanese than Americans, for example. (If you're looking for an MMOG with lots of Japanese you can play with, Final Fantasy XI will interest you.) I remember when I played in the Bomberman Online free open beta, which was Japan only but you could download the client and a patch to make it work with English WinXP, the Japanese were all very nice and respectful. It's unfortunate that the American societal mentality is screw everyone as often as you can to get ahead, but there's nothing you or I can do about it. I can only ask politely that people do not cheat, and if they are suspected of cheating they either prove themselves or risk the banstick. I'd rather not have to look over everyone's shoulder to make sure they're not cheating... :)

this post is a complete generalization of cultures. i myself am asian american, and have spent time in both places. i played everquest 24/7 for 3 years and found no difference in cheating between the american and japanese players on our server. the japanese players are just as big cheaters as the americans. some of the nicest people i met were japanese, some of the nicest people were american. as i've grown older, i find aside from minor difference, the human race is pretty much the same from iowa to indonesia. in any large group of people, you'll have your a$sholes and angels. the way in which their behavior is expressed might be different, they might do it behind your back or in front of your face, either way it's the same. so watch your mouth when you try to slur any country, ESPECIALLY MINE.

as to the issue at hand, i used to care about validating this and doing that so other people could see my accomplishments were clean. well, now i don't care. i overclock for myself, if someone chooses to believe i cheated, i don't really care.

[agenda2005]

This is rather Sad. There's probably no way of showing system stability beyond Super Pi 32M. I will henceforth keep my system stability to myself until the air is clear.

[conrad.maranan]

I have a solution for "Hard to Cheat On" stability testing.

Again, Wizzard, has a solution, and it can be found in his Systool Program.

I tested this program in a few different ways, and I had a hard time cheating on it, see, with it's constant updating the list as to what passed and what didn't, I couldn't create a cheat that looked legit, cause it wouldn't match the next results...
thats very very good, for spotting cheats real fast.

manual cheating, and speed cheating did nothing to this program.

so it's about 90% cheat free, as far as I can tell. :D

http://systool.techpowerup.com
Looks good, Kunaak. But perhaps some type of verification code or online verification will be the only way to be 100% sure. Check this out:

http://www.xtremesystems.org/forums/attachment.php?attachmentid=30635&stc=1

[Kunaak]

haha... well, cheating with photoshop is always gonna be an issue.
at the moment, all I wanted to do, was find a priming method that was atleast very difficult to cheat on.

with Prime 95, theres nothing hard to cheat on.
with SP2004, it's alittle more solid, but still extremely easy to cheat on.

with Systool, and it's constant update to the time, it's very hard to cheat.
but you are right... photoshop skills can easily change that.

[conrad.maranan]

Who knows? Perhaps when benchmarking makes it to the point where it is actually considered to be a sport, application developers will all start implementing verification systems. :D

[Magnj]

I also just disovered the P95 cheat. I was gonna post it but checked here first. Won't say how, but i ended up with -24 hours and -55 minutes of prime stability lol

[typo_pl]

Hi all, this is Johnny Lee, authour of SP2004.

Yes, someone pinged me about the possible 'cheating' of SP2004 reporting the CPU/FSB speeds. While I found a possible solution on the net to calculate the CPU speed directly, which may go into the next release, as I told the emailer, any decent paint program could get around the checks. You'd probably need some screenshot app that was hooked to the net and did a secure handshake with a server to securely grab a SS w/o possible man-in-the-middle intervention.

SP2004 gets the CPU/FSB speed from Windows. See MSDN - WMI - Win32_Processor (http://msdn.microsoft.com/library/default.asp?url=/library/en-us/wmisdk/wmi/win32_processor.asp)

P.S. If you want me to join the conversation, feel free to ping my email address with an obvious non-spam subject line - like coherent english, two or three words would suffice or the title of this thread would have also worked.

J

[agenda2005]

Hi all, this is Johnny Lee, authour of SP2004.

Yes, someone pinged me about the possible 'cheating' of SP2004 reporting the CPU/FSB speeds. While I found a possible solution on the net to calculate the CPU speed directly, which may go into the next release, as I told the emailer, any decent paint program could get around the checks. You'd probably need some screenshot app that was hooked to the net and did a secure handshake with a server to securely grab a SS w/o possible man-in-the-middle intervention.

SP2004 gets the CPU/FSB speed from Windows. See MSDN - WMI - Win32_Processor (http://msdn.microsoft.com/library/default.asp?url=/library/en-us/wmisdk/wmi/win32_processor.asp)

P.S. If you want me to join the conversation, feel free to ping my email address with an obvious non-spam subject line - like coherent english, two or three words would suffice or the title of this thread would have also worked.

J


Johnny,
It's nice to read from you. Thank you for taking your time writting this nice application. I particularly find it useful for determining my 24/7 stable clockspeed. Please the real overclockers need this application badly to be cheat free. I will suggest the following improvement.

(1) Once the FSB and CPU speed is read correctly, you can also set the program to read the CPU/FSB speed every 5-10sec and report the average speed at the end of the run. The reason been that, someone may use clockgen to underclock the speed during prime95 run and then bump it up before the end of the run to show higher overclock

(2) If you can, please try and write the net SS application version for those who are serious and want to show real stability. Any other version will not be acceptable just like super pi mod 1.4

(3) A better timing method that cannot be fundled with in windows, so that time cheat can be eliminated.

(4) An online verification with some sort of checksum that anyone can verify the validity of the run.

Your attention will be appreciated. I've already ping you about all these request as well. just to make it public.
Thanks.

[timpanogos]

What is everyone's take on occt? It claims to use the same concepts of prime95 (constant answer compares). It also appears that when occt fails for me, so does prime, and when it passes, both pass - this is with all other benches working when both of these fail (s&m, super-pi32m, memtest, all the 3dmarks etc.)

The thing about occt is there is no config (fft size, custom, mix, small etc.), one size fits all for a consistent compare.

[AKURA]

this post is a complete generalization of cultures. i myself am asian american, and have spent time in both places. i played everquest 24/7 for 3 years and found no difference in cheating between the american and japanese players on our server. the japanese players are just as big cheaters as the americans. some of the nicest people i met were japanese, some of the nicest people were american. as i've grown older, i find aside from minor difference, the human race is pretty much the same from iowa to indonesia. in any large group of people, you'll have your a$sholes and angels. the way in which their behavior is expressed might be different, they might do it behind your back or in front of your face, either way it's the same. so watch your mouth when you try to slur any country, ESPECIALLY MINE.

as to the issue at hand, i used to care about validating this and doing that so other people could see my accomplishments were clean. well, now i don't care. i overclock for myself, if someone chooses to believe i cheated, i don't really care.

Same here. I stop giving to much :banana::banana::banana::banana::banana: bout validation and all that ***** nonsense. Anyone can cheat as so does speak. Either you'll fall into it and respect the god damn cheater in the end is your priority. Not me and anyone else is. So let's stop doing this cheat spotting and anti cheat program and continue on the legacy from what it started in the beginning. This all a ***** waste of time imo. Everyone here knows who the masters ocing are as far as im concern. :D

Post Edited For Sensor Bypassing.

Attempting to blatently cuss in a post, while mis-spelling the words... don't do that again. Your free to have your own opinion, and voice them as such on the forums, but we do have some basic rules here for good reason. If you can't respect them, then do not post.

-Kunaak-