NVIDIA's shady trick to boost the GeForce 9600 GT

[Scyphe]

Originally Posted by cbunting

Ok, Then let's prove real true stats to see if the article is true.. Anyone else, Post yours!

BFG 9600 GT OC 512MB.
http://www.bfgtech.com/bfgr96512gtoce.aspx

100Mhz PCIe Bus;
http://www.techpowerup.com/gpuz/3h8eb/

115Mhz PCIe Bus;
http://www.techpowerup.com/gpuz/2gu7p/

Oh wait.. Nothing changed.. Notice I didn't use Riva Tuner for statistics.

I used GPU-Z which DOES support the 9600 GT...



Chris

Hahaha, you must be pulling our legs or you have a serious cognitive ability problem.

Did you benchmark your card with 100Mhz PCIe and 115Mhz PCIe?

Did you read the posts that said the drivers do NOT reflect any core changes when overclocking the 9600GT via the PCIe-bus?

This is all like a black hole to you, right? You seriously don't understand a word we're saying, that much is clear.

For myself, I can't. It makes no difference what I OC as nothing changes the clock speeds.

And how do you determine the clock speeds? Since the drivers doesn't reflect the overclocking through increasing the PCIe-bus clock (and thus GPU-Z and any other tool using the drivers to find their numbers) you have to find another way to find out what's going on.

[cbunting]

C'mon man..

Overclocking the PCIe bus as page 2 of the article shows isn't right..

When you overclock the PCIe bus, the only allows the Core Clock of the video card to be OC'd higher. It don't change anything by overclocking the pcie.

Page 2 of that article was concluded based on bus frenqency changes.

http://www.techpowerup.com/reviews/N...9600_GT/2.html

We ran the fillrate test with four difference PCI-E clocks set in the BIOS: 100 MHz, 105 MHz, 110 MHz and 115 MHz.

As you can see, the diagram above is pretty conclusive and shows exactly the behaviour we expected.

My card runs stable at 740Mhz core clock. It gets glitchy above 740Mhz.. Unless I OC my pcie bus, I am not able to push the card higher. With a PCIe bus change to 11-125Mhz, I can oc the card to around 762Mhz core clock. But changes to the pcie bus doesn't change anything in regards to the core clocks other than allowing you to push the card oc a bit higher.

So again, What software provided those stats in the graph? Riva Tuner?

What part of what the Author of Riva Tuner said can't anyone understand?

core clock can be monitored improperly in hardware monitoring module etc.

Furthermore, the 9600 would not even run stable enough at 834 MHz to even finish a benchmark of 3DMark06. If you think it will, then try it yourself.

[Scyphe]

What the hell, which part of the following do you NOT understand? Forget Rivatuner, forget GPU-Z, forget nTune!!!!!!!!!

In order to test our theory we ran 3DMark06 Multitexture Fillrate tests on both the GeForce 9600 GT (G94) and the GeForce 8800 GT (G92). Both use the same underlying GPU architecture, with the G92 being the faster part. We always tested at the stock frequencies of our two cards, we never adjusted the GPU frequency manually.

The 8800GT did NOT gain anything by increasing the PCIe-bus, as expected. The 9600GT DID gain more each time the PCIe-bus was increased, with NO other overclocking going on whatsoever.

Set your PCIe-bus to 100Mhz. Run the 3DMark06 Multitexture Fillrate test. Then set your PCIe-bus to 115Mhz and run the same test again.

[cbunting]

3DMark06, FutureMark nor YouGamers.com support the 9600 GT with any of thier software.

Results Inconclusive?



Wow, YouGamers by FutureMark doesn't even know what card I have.. And to think the outdated 3DMark06 would..

God, I don't even believe in every article.
Chris

[Solaris17]

um they still give you a score.........and when i oc my card the score goes up i know it sounds crazy....but set your pci-e bus at 100mhz run the fill rate post up the numbers...then set it to 115mhz run it again post up the numbers.

[Scyphe]

Originally Posted by cbunting

3DMark06, FutureMark nor YouGamers.com support the 9600 GT with any of thier software.

Results Inconclusive?



Wow, YouGamers by FutureMark doesn't even know what card I have.. And to think the outdated 3DMark06 would..

God, I don't even believe in every article.
Chris

Well, with that you've just exposed yourself as a disinformant. You failed though, nobody seems to have adopted any confusion that you wanted to spread.

[imperialreign]

this thread has become quite hilarious

[Steevo]

^ True.


If I had the money to waste on getting a 9600GT I wouldn't. It has already been proven by multiple respected people what is true, and this joker thinks he can come in with a backseat drivers license and change everyones mind. Fat chance.



Besides, your Pentium D sucks.

[Solaris17]

Originally Posted by Steevo

Besides, your Pentium D sucks.

LOL!

Originally Posted by Steevo

It has already been proven by multiple respected people what is true

thank you

[Wile E]

cbunting:

you are not listening to a single word anybody is saying.

I'll make it very simple to understand:

NO SOFTWARE READS THE CORE CLOCKS OF THE 9600 PROPERLY!!! THE 9600 DRIVER DOESN'T EVEN READ THEM PROPERLY. NO SOFTWARE CAN BE USED TO PROVE THE OVERCLOCK BASED ON CLOCK READINGS. YOU MUST COMPARE PERFORMANCE VIA BENCHMARK SCORE DIFFERENCES AT BOTH 100MHZ PCIE AND THEN AT OCED PCIE SPEEDS.

Now, instead of trying to dig up articles, and trying to show clockspeeds on a program (which is wrong, no matter what program you choose, PERIOD), just do what the hell I said. Run the 3DMark06 benchmark. Not open it to see what it says about your clock speeds, actually run the damn benchmark test. Do this with PCIe at 100MHz, and post your score. Now, do this again, with PCIe set to 110MHz, and post the score. Until then, your words mean absolutely nothing.

[Saakki]

i already earlier told you that cbunting is a hired nVidia mindblender to confuse OC guys and reviewers even more ..why bother buying 9600 GT because its priced almost same as 8800 gt around here...

[MrMilli]

Originally Posted by Saakki

i already earlier told you that cbunting is a hired nVidia mindblender to confuse OC guys and reviewers even more ..why bother buying 9600 GT because its priced almost same as 8800 gt around here...

He's not. He's just plain ol' school stupid, nothing more, nothing less. Ok maybe less.

But seriously, if you read all his posts, you will see that he doesn't know a thing about these stuff. I wonder if he even knows what fill rate means. Maybe his English is just bad ...

[Saakki]

Originally Posted by MrMilli

He's not. He's just plain ol' school stupid, nothing more, nothing less. Ok maybe less.

But seriously, if you read all his posts, you will see that he doesn't know a thing about these stuff. I wonder if he even knows what fill rate means. Maybe his English is just bad ...

hope so..pretty hilarious tough..

[Tatty_One]

Lol, this thread has become my daily read on my laptop at work when I go for a cr*p each morning, in fact it's that good it has taken over from FHM

[OnBoard]

Originally Posted by imperialreign

this thread has become quite hilarious

That was some fun arguing (maybe not for those that took part, but for us readers) I was tired when I started 4 pages back, somehow I'm not anymore

[dtdw]

If thats the case then the fps count would be even higher ?

Nowander my Oblivion's lowest fps jumped from 45(lowest) to 55(lowest)

I didnt dare to O'c my 9600gt in sli cos its stil new but i used Ntune to 'tune' my system so the Pci-e is tune as well and it went up to 2925mhz ..Woot ! ..

Went and test Oblivion and it went 10fps higher...

Now that only means one thing...OC the card till 725mhz, increase Pci-e to 120mhz

So its 725mhz + 130mhz = 855mhz

Edit : Gee now i need to get an active cooling for my southbridge =.= or do i need to? hahahah

Insane

[Wile E]

Well, considering he hasn't responded, I'm gonna guess he ran the tests, saw the increases, and felt like an ass.

[jaydeejohn]

I dont post that often here, but maaaaaan, was that daft or what???? LOL Fellas this has been one great thread, and TY w1z for a great insight into what may be a great deception, especially if I were a writer/reviewer

[Tatty_One]

Originally Posted by Wile E

Well, considering he hasn't responded, I'm gonna guess he ran the tests, saw the increases, and felt like an ass.

Amen to that

[dtdw]

okay i just need to clarify this again, the author manage to check the 'actual' core speed in rivatuner but i cant. Why is that ?

for his example his clock @ 725mhz bt reported @ riva is 783mhz

but whereas if thats the case 650mhz original clock would report 708mhz at riva ...

or did i NOT get the calculation correct..

[Solaris17]

the new version of riva tuner doesnt do that i think unwinder stopped it from reading from the pll

[dtdw]

Originally Posted by Solaris17

the new version of riva tuner doesnt do that i think unwinder stopped it from reading from the pll

Ohhh..so that means i need to use the old version...

or is there a way to edit the config so that it reads the actual 1 from 2.07 ?

heheeh...i am really really curios to know wht is the actual one after raising my pci-e

my card is unstable @ 700mhz while having pci-e @ 117 ... but i thought raising the pci-e 'increases' stability ?

[ghost101]

Originally Posted by dtdw

okay i just need to clarify this again, the author manage to check the 'actual' core speed in rivatuner but i cant. Why is that ?

for his example his clock @ 725mhz bt reported @ riva is 783mhz

but whereas if thats the case 650mhz original clock would report 708mhz at riva ...

or did i NOT get the calculation correct..

rivatuner doesnt report the actual core speed. It assumes a 27mhz multiplicative. So for a 650mhz core with the pci-e bus running at 100mhz, the multiplier is 650/25 = 26.

Therefore, rivatuner should read 27*26 = 702mhz.

If on a pci-e bus of 125mhz, the actual core clock is 26*(125/4) = 812.5. Rivatuner will read, 702mhz still and nvidia will report 650mhz.

[Unwinder]

Originally Posted by Solaris17

the new version of riva tuner doesnt do that i think unwinder stopped it from reading from the pll

Wrong guess. I still read the clocks from the PLL. I just use hardcoded 25MHz crystal clock for G94 based display adapters now.

[AddSub]

Wrong guess. I still read the clocks from the PLL. I just use hardcoded 25MHz crystal clock for G94 based display adapters now.

How honorable. I'm sure nVidia appreciates it.