Onepagebook, coolaler and Superpi

[massman]

http://www.xtremesystems.org/forums/...d.php?t=126554

His results are blocked at hwbot, I'm waiting on Rich' call for further actions.

[M.Beier]

Lower clocks makes that number lower, right ? Or is it linear @whatever clocks ?

Something to see this easy is to make at low speeds so everyone can make the tests.

@3500mhz I got 14,297secs with fat body @583 4-3-2-1 ( 2x512 ) which gaves me a nice ~50000 points. I know that it could be a bit better, not too much tho.

OPB made a run of 14,15secs @3,5ghz with 440 3-3-3 ram ( 2x1gb )

That made me think of it... Also one user which gets 14,25secs with an gigabyte 965p and ram similar to OPB.

What do you say ?

A bit yes..
@356x and 13.9x, with less then 500mhz mem at 4-3-2-4..

[Raja@ASUS]

this thread has no concrete proof. I can say that OPB is one of the most honest people I know and does not hide his settings or tweaks from anyone...

In fact he'd prolly invite you round to his place to repeat those scores...

Bad thread, should be deleted..

sorry to see this sort of stuff here, what a shame.

[M.Beier]

Lets see them then.

Read over the graphs and post carefully everyone. This is not about opinion here.

Gautam, search, Im pretty sure I showed around that times in that "Step5 vs. step6 thread" you do remember that one
Honestly, I've been movin' away from home, havnt spend much time playin' with the rig, but just purchased new gear, so I cant really boot the primary rig.. (wont be up for hmm, lets say, atleast next wedensday.. - Party and family events all the way 'till monday, and then two days of hard study..)

EDIT:
Picture not in thread, try to search on http://tweak.dk @ "SPI1M bedste tider", should pop up
- And yes, Im on that list a few times, hihi

[Gautam]

Actually, its possible to break 50000 at speeds well below 4 GHz, but here we're mostly dealing with 4.5GHz+ scores.

The 250+ data points in the graph speak for themselves.

Like I said in my now quoted and deleted post, everyone please read the first post and examine the graphs objectively before rushing to post.

[Jort]

that's what i mean, the point where instability somehow is stable..

Who else hasn't got the same experience,

you got run 2970mhz superpi al day long, 3GHZ would error, but after idling and pushing it again somehow it manages to work just that once, and give you that little nice extra. i have multiple times on A64...

juts brainstorming eh before you guys ramp me over lol :p

Yes i had this couple times on A64 too.

Never could get the same time again on the same settings, so is that score invalid? or cheating? No

[Lithan]

I'm not sure I get what pall and jort are saying. Are you saying that you believe the benches aren't representative of the speeds they actually were run at because the run was done, then the system was downclocked, then the screenshots were made?

[Dumo]

if i can run at that speed

QFT. Is Not easy.

[Gautam]

I'm not sure I get what pall and jort are saying. Are you saying that you believe the benches aren't representative of the speeds they actually were run at because the run was done, then the system was downclocked, then the screenshots were made?

They're saying that if the system is unstable enough, you can get a result that's unusually fast at times.

[alexio]

I'm not sure I get what pall and jort are saying. Are you saying that you believe the benches aren't representative of the speeds they actually were run at because the run was done, then the system was downclocked, then the screenshots were made?

They are saying the error check or internal clock is failing due to instability resulting in non legit times.

[M.Beier]

Gautam, sorry, I misread your post I guess.. I just thought you accused me of not truthin', and that I do not like (would be second time this evning as well.. Kunaak just did in another thread about RAM.. GAH)

I couldnt break it at 4ghz+, due mobo failed with good mem settings :'(
- I belive SIOUX has tho'... He beated my score swiftly after, with lower clocks, but higher mem

[Chicco85]

Whatever is the reason of the stunning scores (in the positive or negative sense of the term), i believe that Vapor's work is very interesting.

I like the overall analisys of the data. Great work.

[Gautam]

Gautam, sorry, I misread your post I guess.. I just thought you accused me of not truthin', and that I do not like (would be second time this evning as well.. Kunaak just did in another thread about RAM.. GAH)

I couldnt break it at 4ghz+, due mobo failed with good mem settings :'(
- I belive SIOUX has tho'... He beated my score swiftly after, with lower clocks, but higher mem

As of right now the best at 4+ GHz result on hand is kyosen at 4167MHz, 12.047 resulting in 50200.

[zakelwe]

this thread has no concrete proof. I can say that OPB is one of the most honest people I know and does not hide his settings or tweaks from anyone...

In fact he'd prolly invite you round to his place to repeat those scores...

Bad thread, should be deleted..

sorry to see this sort of stuff here, what a shame.

No concrete proof ? Sorry, there are some graphs on the first page that took a lot of time and effort that show an abnomality.

Regards

Andy

[Gautam]

Whatever is the reason of the stunning scores (in the positive or negative sense of the term), i believe that Vapor's work is very interesting.

I do too, and am glad to see someone else agrees at least on that point.

[Dumo]

It will be a good idea for any of us to run 6800 at the same 5.5Gish and the same ram timings and compare it to the results.

[zbogorgon]

It will be a good idea for any of us to run 6800 at the same 5.5Gish and the same ram timings and compare it to the results.

you have picoo's superpi, 5600MHz, same mobo, P5WDH, same timings, 3-3-3-1, 9.013s superpi

that is 50MHz more and 0.110s less

[Gautam]

Once again, please refer to the first post. There are a handful of 5.5+ results.

And an additional one as of today, hipro at 5514 resulting in a time of 9.219s. His performance product is 50833.

I might remind everyone that hipro has a strong history of out-tweaking everyone in SuperPI. His Dothan times were untouchable, even by some Team Japan members at a variety of different clock speeds.

Other 5.5GHz+ times:

9.016 5620MHz 50660, Picoo
9.093 5568MHz 50629, MemoryTeam Italy
9.330 5506MHz 51370, maxxpct

[Pedro Rocha]

This issue have been discussed amongst moderators for days, we only decide to going public have we have strong evidence that OPB and Cooaler's results in superpi are abnormally good for the CPU speed..

Some sort of explanation is due either it's a bug, a cheat ot or a tweak. All XS comunity will deserve a clear explanation.

[^don.k's^]

A bit yes..
@356x and 13.9x, with less then 500mhz mem at 4-3-2-4..

Those 60 extra mhz's make that time slightly better than average. I meant 3500 plain mhz's, not 35xx.

[NiCKE^]

So this is actually a desperate attempt to either get coolaler and opb marked as cheaters or getting tweaks that you guys ain't aware of that they don't want to share with you?
Do you really think they will share stuff with you when you do like this?

So OPB's score isn't valid? Well the validation check says it's.

So when someone is a better tweaker then avarage joe it has to be some bad stuff involved?
OPB has been sharing tweaks and helped your members here at XS since he first came here. Now you treat him like he is a bad guy. Oh yeah XS staff we know you don't like OPB and OCX, but this smells like jealousy.

[boblemagnifique]

That make just 1 year which one speaks that scores are impossible on be half of OPB, but good I passed for a idiot here, That give pleasure that some reflect a little instead of believing somebody who benefits from the scores has reasons professional .....

http://www.biwa.ne.jp/~yok/28GP/28GP.htm

Benchs and overlocker are a passion, one should not mixed work and the pleasure

[cpulloverclock]

My best time-frequency ratio (I have a 11.9s at 4203MHz)

Vapor, look at my best spi time (time-frequency ratio) :

0'17"000 at 2928MHz
http://img186.imageshack.us/img186/9135/17000vq8.jpg

0'13"828 at 3603.2MHz
http://img180.imageshack.us/img180/711/13828sm4.jpg

0'13"484 at 3699.0MHz
http://img282.imageshack.us/img282/2...4843699vu6.jpg

0'12"078 at 4149MHz
http://img187.imageshack.us/img187/520/12078gu7.jpg


So you see my best ratio is at 2928MHz (49642), when I see OPB to get 494XX at 5550MHz --> [I'd say 5650MHz], I laugh cuz at 3GHz he would have around 48800...

No comment

PS : I'm ok for his 8"906 of course (his score is correct), no problem for that but for the frequency... I'm sure but it's personal

[Gautam]

So this is actually a desperate attempt to either get coolaler and opb marked as cheaters or getting tweaks that you guys ain't aware of that they don't want to share with you?
Do you really think they will share stuff with you when you do like this?

So OPB's score isn't valid? Well the validation check says it's.

So when someone is a better tweaker then avarage joe it has to be some bad stuff involved?
OPB has been sharing tweaks and helped your members here at XS since he first came here. Now you treat him like he is a bad guy. Oh yeah XS staff we know you don't like OPB and OCX, but this smells like jealousy.

No one accused anyone of cheating. All that was called for was an explanation.

Personally, I don't feel that anyone is above question. The deviation shown on the graphs is large enough that some questions certainly are warranted.

[zbogorgon]

it is suppose to be for freqs 4500+, not low freqs