Onepagebook, coolaler and Superpi

[Lithan]

It's OFFENSIVE to them......So PLEASE FORGET this word......

It's a shame that what should be a useful abbreviation for japanese was used as a slur (from WW2 on?) but that's the case, so yeah. Saying it is no good.


That said... I havent even bothered looking at system clock in bios for years (is it even still there?) but windows USED to be able to tell system clock what was what... so yeah. Unless that has changed, if you can hack windows clock, you can hack system clock. It wont be realtime, but it will be close enough (sub1ms) that no x.xxxxs could tell any difference. But I don't know how cpu-z works... I cant think of anyway to prevent that kind of hack, but I'm no expert on the subject by any means. I'm just a guy who would forget to adjust bios clock and wind up doing it via windows.

LoL this is starting to remind me of the time I noticed that if I cleared cmos, loaded windows and started prime, then set clock I could prime for several years in a matter of seconds. :P

[Shon]

Well it's 10+ years you can freeze an app and read/modify whatever var you want inside it, just need the right app.
Just put back a little bit the internal clock every step and you got a perfect wr really hard to catch.

[Langley]

Oh oh oh oh oh... Merry Christmas... and that's about 5 months that i talk for nothing....

[Omastar]

PallMall's avatar confuses me and quite frankly I am tired!

[zakelwe]

aye but 06 is so cpu bound in SLI, you just need watercooled gpus and prolly a 5.2GHz QX6700, you'll have insane GT1/2 scores to make up. Not greedy i just want 5.1GHz Kentsfield
so let's say i manipulate to 5.1GHz, will it run like a 5.1GHz processor?

If the mechanical judges are not up to the job then it will be up to the humans to catch them out like you did before.

At least with 3dmark we have 4 benches to judge on and someone putting up a score in one is likely to have to put up a score in some of the others and then will maybe make a mistake in how the cpu / gpu dependency on different runs and games differs. Maybe it is too difficult compared to one simple SuperPi run ?

Since the post earlier where you caught out Coolaler with the 6800U how many times have you, as a top 3d overclocker, caught someone out ?

Regards

Andy

[uclajd]

Exactly. People make millions of secure transactions on the Web every day via SSL. There has got to be a way to do this. All these supposedly Xtreme people here you'd think someone would want to step u and be an Xtreme Legend like Franck and make an invaluable piece of software.

Is this not a huge reason to get a community effort for a simple benching program (a la Super Pi) that is developed by clockers, for clockers, and possibly even open source.
I'm sure there are more than enough programmers here to be able to make a secure program that can let us alleviate these kinds of threads.

I won't even get into "blah is guilty" or "hoo-ha seems dodgy" but I will state that a lot of the tweaks I've learnt from this forum have come from OPB and for that I am grateful.

[Vapor]

aye but 06 is so cpu bound in SLI, you just need watercooled gpus and prolly a 5.2GHz QX6700, you'll have insane GT1/2 scores to make up. Not greedy i just want 5.1GHz Kentsfield
so let's say i manipulate to 5.1GHz, will it run like a 5.1GHz processor?

As Andy said, we have more benches to base judgement off of, and can compare them as well.

If you manipulate to 5.1GHz, from say 3.4GHz, everything will run exactly 50% faster. Even more GPU-bound tests. In the case of a 3.4GHz QX two well cooled/clocked 8800GTXs, it will be easy to spot because the GPU-bound tests will be much too high.

[aicjofs]

If I learned anything from these 22 pages, is that it's a good way to waste the morning...

One thing that stands out I don't understand and people keep saying, I don't think we need "to have the tweak"(if it exsists). There are other ways to provide proof that it's a tweak not a cheat without surrendering the tweak.

[uOpt]

Here is two OPB scores viewed in this thread, lots of differences, and the last loop is strangely different compared to other, not the same on the 12.734s.

These are artifacts that can come from JPEG compression. Different software, different original image and different quality levels all cause artifacts.

Can we please close this silly discussion about the grey pixels?

You will NEVER be able to take a digital image and be sure it is not a fake. This is pointless.

[Vapor]

These are artifacts that can come from JPEG compression. Different software, different original image and different quality levels all cause artifacts.

Can we please close this silly discussion about the grey pixels?

You will NEVER be able to take a digital image and be sure it is not a fake. This is pointless.

I completely agree. Heck, even same software, same compression algorithm, but putting the text up or down a few pixels relative to the compression grid can cause different artifacts.

Please drop the photoshop cheat talk....there's literally no way to prove or support it. If a cheat were to happen, it would likely be with a speed hack anyway.

[Lithan]

Everyone knows that only I have the photoshop skills to pull off a hack like that anyway.

[itznfb]

lol, you didn't even bother to take out the little pieces left from the red circle

[Vapor]

Please stop the photoshop talk....that does nothing for the thread.

PallMall, that's an old question....does it mean anything? IDK, but if you look at loops 18 and down, the numbers are okay....seems to be a small bug/instability.

[massman]

Ok for me it's clear now, Coolalers results is fishy!
the guy get's beaten in LOOP 19 but manages to get a miraculous flat 9.000secs?
Also every jump from the lats loop to the final loop is at least teh same as previous loops or higher then 0.3secs, this one is only 0.140

Coolalers loop 19 is indeed impressive. I compared it to Hipro's latest run: Coolaler was behind on loop19, but managed to do 9s flat, while hipro got 9.022

The initial value and output need to be very good to have a WR, basicly, those are the two that you need to tweak (as said by OPB once)

[zakelwe]

Strange bug lol

Yes I agree, when things have been bought into question it is best to question everything even if it does not spread light on the matter in hand as it might point to something else

However, we have not had any answer to the graph on the first page never mind the subsequent AMD28GP abnormality so peculiarites with the last loop of SuperPi will have to stay down the pecking order I am afraid.

To be honest this thread I feel is about to come to a close. Lots of good info has been posted and you can make your minds up one way or another.

Regards
Andy