狂少's CDT-tweak

[Onepagebook]

This is the best I've ever done at 3600MHz Kevin.

sorry about del your screeny

If you're willing to divulge whatever it is that you're doing leading to such fast times, I, and surely everyone else here are all ears.

I think I am done with that. I will let fugger knows that I just could not do as what I and him said which is about what we think the best way to do,

I have been doing my best try to explain and everythiny turns like I am like . so ya, I think I am done for that.

[Gautam]

Okay, as you wish. Is there anything else you'd like to see in this discussion?

[Onepagebook]

OPB, 49500pp at 3600 and with <600MHz DDR2 just isn't possible with Conroe/Kentsfield....heck, it's BARELY possible with DDR3 clocked sky-high.

Not only that, but here's a 48444pp you tried to pass off as the 36GP WR before:

That ^^^ is frankly impossible...

This tweak may be worth a few seconds on 32M, which is great...thank you for sharing. However, let's call a spade a spade and acknowledge what scores are impossible at the claimed clocks.

did you ask my permission to use that screeny? I mean not post , but you copy the picture and re-upload to your space, is that considered as stealing?

[Charles Wirth]

Kevin thanks for posting.

Everyone else,
Please keep it civil, I want to see this worked out. I have been in orientation for 8 hours today and I cannot think straight atm. I will be home tomorrow to read up as I need to catch up.

Thanks

[Onepagebook]

Kevin thanks for posting.

Everyone else,
Please keep it civil, I want to see this worked out. I have been in orientation for 8 hours today and I cannot think straight atm. I will be home tomorrow to read up as I need to catch up.

Thanks

Certainly Sir, that's what I promise you lastnight, however, not what I would like to say thinkgs beyond my business. I just don't want your great co-admin jump up from nowhere and point the finger to me like that which is lot more different than what we discuss on the phone last night.

I highly appreciate that your input here and take care everything.

[Vapor]

did you ask my permission to use that screeny? I mean not post , but you copy the picture and re-upload to your space, is that considered as stealing?

I did not upload/transfer it...member Final1977 has posted exactly that on XS in a thread that's since been removed. He was wondering if it was legit. There was also a ludicrous 32M screenshot in the same post.

Back to my original point....you're not a victim here, for the longest time you've posted screenshots of SuperPi runs that are just not possible. When asked how you did it, or if you can back it up, or ANYTHING like that, you throw a hissy fit (like now).

This CDT tweak is worth a few seconds in 32M, maybe .016s in 1M...it does not 'clear your name' in the slightest since it's frankly not good enough to account for the performance boost required to achieve your 'WR' scores.

[Onepagebook]

I did not upload/transfer it...member Final1977 has posted exactly that on XS in a thread that's since been removed. He was wondering if it was legit. There was also a ludicrous 32M screenshot in the same post.

Back to my original point....you're not a victim here, for the longest time you've posted screenshots of SuperPi runs that are just not possible. When asked how you did it, or if you can back it up, or ANYTHING like that, you throw a hissy fit (like now).

This CDT tweak is worth a few seconds in 32M, maybe .016s in 1M...it does not 'clear your name' in the slightest since it's frankly not good enough to account for the performance boost required to achieve your 'WR' scores.

oh then I feel bad for you if you think that;s impossible or whatever. just like why OSU can beat umich...sad, just take the loss
sounds like back to 199x year...what did umich said.." there is no possibility that OSU can beat umich, cuz history said so , then don't even think about 3 in a row by same quarterback.." well then what happened? you ARE SO pathetic, buddy

[Vapor]

oh then I feel bad for you if you think that;s impossible or whatever.

Then prove it's possible...that's what we're ALL waiting for. Your scores are so far out of line that it's one of two things: a cheat or a massive tweak. At this point, you have too many scores WAY out of line to be a 'bug.'

FWIW, you have yet to show us a name-clearing massive tweak (CDT is not massive...at all) and it's been a year....that's why everyone is defaulting on you having cheated. Your results, your defense, and your reactions speak volumes about how these scores were achieved.

[Onepagebook]

Then prove it's possible...that's what we're ALL waiting for. Your scores are so far out of line that it's one of two things: a cheat or a massive tweak. At this point, you have too many scores WAY out of line to be a 'bug.'

FWIW, you have yet to show us a name-clearing massive tweak (CDT is not massive...at all) and it's been a year....that's why everyone is defaulting on you having cheated. Your results, your defense, and your reactions speak volumes about how these scores were achieved.

oh I donn't need to if this is just your own trick to lead me show my tweak, sorry and ya, I forgot to tell you. if people are think they wanna know what and how I have been study in this charming calculation, they will go to OCX and discuss it. because they are sick of the person like you and of course some fanboy-ism here. XS is NOT like ..let say 5 years ago? what's the reason behind this...ROFL

[G H Z]

It seems to me that some recent posts in this thread violate forum rules, ie flaming. Can someone please take out the trash?

[s7e9h3n]

Everyone here should just take a deep breath and stop taking everything so seriously. In OPB's defense, he shared one of his tweaks which benefits him in his SPI runs. They may or may not help everyone, but seemingly it's effective for some. This is a discussion about the validity of this tweak, not a debate regarding OPB's history. If the tweak helps some people here, then great. If it doesn't, then so-be-it. TBH, if it was anybody besides OPB who offered this tweak, people would be applauding the person's attempt rather than criticizing his character.....

BTW, I'm not sure if everyone's aware of this, but OPB owns an RAM EEPROM programmer which allows him to change almost all of the 100+ subtimings which most of us will never see. This could partially explain why some of his runs seem unusually fast for such relaxed primary timings.

[Vapor]

Ugh, I'm not trying to trick you Kev. I don't really care what the tweak is, but at this point, it's your last bastion of hope to clear your name....you REALLY should care. At least try to publicly prove your scores are legit...you've managed to talk your way out of backing up your out-of-line scores 100&#37; of the time.

No one at OCX can even come close to replicating your scores either...no one anywhere in the world can. And every time someone challenges your clock-for-clock performance, you do exactly what you're doing now...trying to talk your way out of it with no real effort to truly vindicate yourself.

If you actually do have a tweak that can clear your name, why don't you share it with FUGGER and have him replicate your performance products. Or s7e9h3n (who unfortunately doesn't run Intel )...or k|ngp|n, or Hipro, or any of Team Japan...or any other legend, one that you feel comfortable with.

[Onepagebook]

Ugh, I'm not trying to trick you Kev. I don't really care what the tweak is, but at this point, it's your last bastion of hope to clear your name....you REALLY should care. At least try to publicly prove your scores are legit...you've managed to talk your way out of backing up your out-of-line scores 100% of the time.

No one at OCX can even come close to replicating your scores either...no one anywhere in the world can. And every time someone challenges your clock-for-clock performance, you do exactly what you're doing now...trying to talk your way out of it with no real effort to truly vindicate yourself.

If you actually do have a tweak that can clear your name, why don't you share it with FUGGER and have him replicate your performance products. Or s7e9h3n (who unfortunately doesn't run Intel )...or k|ngp|n, or Hipro, or any of Team Japan...or any other legend, one that you feel comfortable with.

why I don't because you guys see team JP as a Team JP no doubt and whatever they post is priceless, BUT you guys see me as garbage..so that's why I don't want even give a try to post @ here. you see what S7e9h3n try to talk? you just don't get it do you?
no one can achieve? so you think elmor is a cheater as well? what a blind admin here

[Gautam]

elmor is an extremely skilled bencher who has my highest degree of respect. I do not see how that is relevant.

Team Japan has a couple of virtues on their side...modesty, respectfulness, and a willingness to answer any question that is asked of them as thoroughly as possible.

[hipro5]

BTW, I'm not sure if everyone's aware of this, but OPB owns an RAM EEPROM programmer which allows him to change almost all of the 100+ subtimings which most of us will never see. This could partially explain why some of his runs seem unusually fast for such relaxed primary timings.

I got one too....so what?.....

I will not "debate" in this conversation - for the time been - I will only read.....

[Vapor]

OPB, no one is questioning any plausible results. Team Japan pushes the limits of clockspeed and overall hardware speed and achieves their results with amazing clockspeed and tried-and-true tweaking we've all learned from them.

Yours are beyond implausible, they're impossible. It's nothing about you, it's your scores, your defense, and your reactions.

Speaking of defense and reactions...Gautam has a good point about the way Team Japan conducts themselves. They're extraordinarily helpful and willing to share what they know and have honor in their scores and their hard work. We don't see this from you...plain and simple.

s7e9h3n has a good point...we are off topic, this thread was a thread about a good tweak that you came up with. Though the fact that you came back to XS and in this thread posted "I won't be back here and post until my name is cleared up here in public." and expect CDT-IV to clear your name brought all this up. While it's a good tweak, it's not nearly that good. And only you can clear your name...it doesn't happen on its own.

And no, subtimings will not change SuperPi times nearly enough to compensate for your totally out-of-line scores. Heck, primary timings don't even do THAT much.

elmor has not come close to replicating your 'WR' 1M times/performance products from a year ago, or the 49500 in this thread, or the 48444pp in this thread....no one has. He is a great bencher and deserves our respect, but frankly, he has nothing to do with this. Your scores are really and truly out of line and you're the only one who knows how they happened....

Again, this is an opportunity for you to try to make an effort to prove your scores are legit....

[KiD0M4N]

Though I risk getting banned for saying this, but I don't understand one thing. Is it "required" to post how you got a good score to be consider legit? I mean, its his tweak, his score... and common, scores keep falling every other day.

Consider this tweak in question. He came up with it right? No one before that thought of something like this... but he did. Why do we assume he stopped there? And why should he give away his prize tweaks in public when people are flaming him left/right/and center.

I believe, in all entirety, that him posting this tweak proves what he is capable. Forcing him to divulge other tweaks to "clear" his name is just not done. How do you know he doesn't have 10 such tricks which in their own self give around 0.1->0.05s advantage, but when done together by a pro, would give the advantage we see in OPB's score.

OPB, all respect to you.

OPB, no one is questioning any plausible results. Team Japan pushes the limits of clockspeed and overall hardware speed and achieves their results with amazing clockspeed and tried-and-true tweaking we've all learned from them.

Yours are beyond implausible, they're impossible. It's nothing about you, it's your scores, your defense, and your reactions.

Speaking of defense and reactions...Gautam has a good point about the way Team Japan conducts themselves. They're extraordinarily helpful and willing to share what they know and have honor in their scores and their hard work. We don't see this from you...plain and simple.

s7e9h3n has a good point...we are off topic, this thread was a thread about a good tweak that you came up with. Though the fact that you came back to XS and in this thread posted "I won't be back here and post until my name is cleared up here in public." and expect CDT-IV to clear your name brought all this up. While it's a good tweak, it's not nearly that good. And only you can clear your name...it doesn't happen on its own.

And no, subtimings will not change SuperPi times nearly enough to compensate for your totally out-of-line scores. Heck, primary timings don't even do THAT much.

elmor has not come close to replicating your 'WR' 1M times/performance products from a year ago, or the 49500 in this thread, or the 48444pp in this thread....no one has. He is a great bencher and deserves our respect, but frankly, he has nothing to do with this. Your scores are really and truly out of line and you're the only one who knows how they happened....

Again, this is an opportunity for you to try to make an effort to prove your scores are legit....

[Gautam]

You're not risking being banned.

The issue with OPB's scores in particular is that they are statistically inconsistent not only with other scores but also with themselves. By which I mean, it takes others several hundred MHz higher to match him. Such gains are too large too be readily explained by any sort of tweaking. However at the very least we wish to get some sort of confirmation that it is indeed a tweak or a group of tweaks that are responsible...and even that we haven't received yet.

In fact, any request for an explanation is usually met with nothing but flames on OPB's part.

[Nosfer@tu]

Well benching depends on What you can make your hardware run.

Lets take a theory, I make ALL my hardware run at the same speeds as KINC or Hipro5. My Scores just cant mach theirs.

I suck at benching, But when you acheave the same speeds other people or even Kinc and Hipr05 will help you.
they migth not give you ALL the tweaks, but at somepoint they will get feed up with being faster than you just because of a tweak..

Ergo if you are faster because of a secret tweak it migth as well be cheating.
We all know it is possible to cheat. and alot has been busted cheating.

But it is not cheating, it is just a tweak yes ?
I dont think so, but I would be borred if I was the fastest in the world, NOT because of my skill with benching and geeting my hardware to run the fastest. But because of a tweak

[Vapor]

Though I risk getting banned for saying this, but I don't understand one thing. Is it "required" to post how you got a good score to be consider legit? I mean, its his tweak, his score... and common, scores keep falling every other day.

Consider this tweak in question. He came up with it right? No one before that thought of something like this... but he did. Why do we assume he stopped there? And why should he give away his prize tweaks in public when people are flaming him left/right/and center.

I believe, in all entirety, that him posting this tweak proves what he is capable. Forcing him to divulge other tweaks to "clear" his name is just not done. How do you know he doesn't have 10 such tricks which in their own self give around 0.1->0.05s advantage, but when done together by a pro, would give the advantage we see in OPB's score.

OPB, all respect to you.

These scores are very out-of-line. That is a fact...

When a score is so, SO far out of line...on a benchmark that SO many people run and get extremely predictable scores based on hardware settings...it raises a few serious questions. The fact that this is OPB means that the scores weren't immediately dismissed. Almost anybody else, and they would have been immediately dismissed and blackballed from the community. We gave OPB his time.

In fact, it's been a year.

In fact, he never once even made claim that his scores were so great clock-for-clock, perhaps not realizing it. Even now, calling the performance product lousy, it doesn't instill confidence that he understands what the performance product can mathematically show about a score. It shows that his scores are great. So great that no one else has come within a mile of achieving them.

Let's repeat this: his scores are seriously out-of-line with ALL existing scores. No scores that are so far above what others have done can be accepted at face value, that'd be a disservice to the OC community. Regardless of who achieves them.

Not only have we not learned of a tweak (even after a year) that can come close to giving this kind of performance, we haven't even gotten a hint of one. We've gotten nothing but hissy fits, topic changes, and flames from OPB when asked for an explanation.

In all respect, KiD0M4N, but this CDT-IV tweak is nowhere near effective enough to account for but a small fraction of OPB's performance advantage. This tweak is worth maybe 10-15MHz over a well-tweaked system and his performance is hundreds of MHz ahead. This is not something we can just 'accept.'

[KiD0M4N]

You're not risking being banned.

The issue with OPB's scores in particular is that they are statistically inconsistent not only with other scores but also with themselves. By which I mean, it takes others several hundred MHz higher to match him. Such gains are too large too be readily explained by any sort of tweaking. However at the very least we wish to get some sort of confirmation that it is indeed a tweak or a group of tweaks that are responsible...and even that we haven't received yet.

In fact, any request for an explanation is usually met with nothing but flames on OPB's part.

Hi Gautam,

Thats cool to know

I agree, some sort of nod or acknowledgment from OPB's side that these are indeed tweaks (hardware software, etc) which gives him the advantage would make everyone a lot more comfortable. I agree, statistically, his "performance level" falls higher (in the opposite sense) than others... but thats how the rule of percentile works right? The top rankers in any IQ test are usually a grade above the rest of the crowd. That kind of a high score doesn't make their score invalid... just that they are better at what they do (in this case Spi benching) that the rest of the people.

Also, consider that FUGGER wants this matter cleared up... he must be convinced there is truth in OPB and therefore he still converses with him and tries to get him back on the forum. I am sure he didn't want to invite him to continue the flaming.

Also sometimes I see a slight bias here... the tone of some people suggests that they simply cannot digest that OPB can get such good scores... they will say anything, go to any extent to discredit him just for the heck of it.

All said and done, at 13.4 from < 3.6GHZ!!!!! Best I have done is a 13.89s with 3-3-2 timings on 975x on a 4MB C2D!

[Onepagebook]

You're not risking being banned.

The issue with OPB's scores in particular is that they are statistically inconsistent not only with other scores but also with themselves. By which I mean, it takes others several hundred MHz higher to match him. Such gains are too large too be readily explained by any sort of tweaking. However at the very least we wish to get some sort of confirmation that it is indeed a tweak or a group of tweaks that are responsible...and even that we haven't received yet.

In fact, any request for an explanation is usually met with nothing but flames on OPB's part.

if you can understand this, I don't see why I can't, if I am a cheater with high definition..lol, would not I can't cheat with more skillful or even cheat it closer and cheat it more legit in the range of Vapor's stupid formula? so I can let the score inside of the legit range and I got WR as well
so that's why I said the whole thing is being set up wake up people

oh I remember about somethinglike when barry Bonds or Jordan...(of course I am NOT comparing I am like them, I am not a smart guy , far more than that) achieve some amazing milestone, they always were doubt if they take any drug..oh ya, now I understand... and is there anyone can even touch them, nope.

[TiN_]

Having

RAM EEPROM programmer

means nothing. It can't get memory chips work better than they can.
I have programmer too, EVERYBODY in this thread and XS in general can assemble it at home, it need only 5-7 cheapo parts

http://tin.topmods.net/spdprog.jpg - Mine one, with DDR1 connector, but this is no matter, SPD chips are equal, in DDR1 and DDR2 and DDR3.

[hipro5]

ON TOPIC......Has anyone else tested the "tweak" of this thread yet?...

After some testing of yours, I will give you a tweak that it'll produse better resaults on your machines.....BUT don't expect to be out of "limits/line"....

[KiD0M4N]

I agree, if score were accepted on face value, then people would start getting away with a lot. However, where do we draw the line? Imagine if people started bantering all of kingpin's threads saying that "dude, this score isn't legit! show me how you did this, or else, you suckz!" Why should kingpin take this bum rap, he will cleanly show the door to the fellow.

I am sure I am not even aware of 1% of all that has happened over the year with OPB because I am sure a hell lot must hv gone down in XIP forums before this thread was allowed in public (just a wild guess.)

@Nosfer@tu: Again, we need to draw a line on whats a tweak... and whats a cheat. It is not something you can judge by following a rule book, but rather by conscience. Is it something you can do openly infront of others? Is it something someone can discover given the time and effort? Thats clean by my book.

Ok, I know I am totally n00b in this, how do you categorize the LOD tweak in 3d benchmarks? It makes the GFX look like right, but gives good scores at the same time. (Please correct me if I am wrong.) Isn't that cheating? Well again I consider it a tweak because everyone can do it and optimize it given time and effort.

Coding a small background to fudge with the system timer to get lower scores etc, well thats BLATANT cheating.

Remember, just like the different shades of gray, the extent of a tweak->cheat is very blurred.

These scores are very out-of-line. That is a fact...

When a score is so, SO far out of line...on a benchmark that SO many people run and get extremely predictable scores based on hardware settings...it raises a few serious questions. The fact that this is OPB means that the scores weren't immediately dismissed. Almost anybody else, and they would have been immediately dismissed and blackballed from the community. We gave OPB his time.

In fact, it's been a year.

In fact, he never once even made claim that his scores were so great clock-for-clock, perhaps not realizing it. Even now, calling the performance product lousy, it doesn't instill confidence that he understands what the performance product can mathematically show about a score. It shows that his scores are great. So great that no one else has come within a mile of achieving them.

Let's repeat this: his scores are seriously out-of-line with ALL existing scores. No scores that are so far above what others have done can be accepted at face value, that'd be a disservice to the OC community. Regardless of who achieves them.

Not only have we not learned of a tweak (even after a year) that can come close to giving this kind of performance, we haven't even gotten a hint of one. We've gotten nothing but hissy fits, topic changes, and flames from OPB when asked for an explanation.

In all respect, KiD0M4N, but this CDT-IV tweak is nowhere near effective enough to account for but a small fraction of OPB's performance advantage. This tweak is worth maybe 10-15MHz over a well-tweaked system and his performance is hundreds of MHz ahead. This is not something we can just 'accept.'