NVIDIA's shady trick to boost the GeForce 9600GT

[Saakki]

thanks for warm welcome Solaris + for the info..my ye olde AGP rig is gettin a bit cheesy

 

[hat]

i think Cbunting is a shady nVidia mindblender..after this..and hello all im new face in your nice forum..after this..should i go 9600 GT or 8800 GT for my upcoming rig..?

THE AVATAR... IT BURNS!! =/

 

[Saakki]

no..it stares at u..BEWARE

 

[trt740]

now Wizzy whats the deal with the 8800gs

a 27 mhz crystal is already on the board for the memory clock, so why not use that like we did for the last 25 or so years?

I can tell it should go higher any news why nvidia driver locked it the Gpu won't go much higher than 700Mgz when tested with with ati artifacting tool. I see people reporting higher clock but they just aren't stressing there cards hard enough.

 

[Tatty_Two]

I can tell it should go higher any news why nvidia driver locked it the Gpu won't go much higher than 700Mgz when tested with with ati artifacting tool. I see people reporting higher clock but they just aren't stressing there cards hard enough.

NVidia didnt want it to compete with the 8800GT....or 9600GT????

 

[Saakki]

lold i misread earlier...ofc 9600 wont beat 8800 gt ...it was bout gs..appologies

 

[trt740]

NVidia didnt want it to compete with the 8800GT....or 9600GT????

I can tell just by overclocking it there is no way it should artifact at 705 mghz or anywhere near that. It's max temps in my system under load, with a dual or are 43c. With a 8800gt, which uses basically the same Gpu, you would get 780/825mghz and this card has less shader=less heat so it should go higher. I'm scratching my head here. I cannot believe someone hasn't unlocked it , it reminds me when ATi first did it to the 3870 cards. My memory won't go higher but this GPU should big time.

 

[trt740]

lold i misread earlier...ofc 9600 wont beat 8800 gt ...it was bout gs..appologies

not by much, and only because the 8800gs is bios locked or driver locked.

 

[Tatty_Two]

I can tell just by overclocking it there is no way it should artifact at 705 mghz or anywhere near that. it's max temps in my system under load, with a dual or are 43c. With a 8800gt which uses basically the same Gpu you would get 780/825mghz and this card has less shader=less heat so it should go higher. I'm scratching my head here. I cannot believe someone hasn't unlocked it it reminds me when ATi first did it to the 3870 cards.

Unfortunatly, you cannot unlock what is laser cut, sadly I think the days of simply disabling hardware have well and truly gone. The next thing you need to look at is voltage, they may have undervolted the card, dump the bios and have a look at it in Nibitor, you should be able to read it at least but of course Nibitor won't let you write to the BIOS juust yet as there is no support for the GS yet.
Then google around for what the boards max voltage is....I am willing to guess that the boards can take/give a lot more than is regulated in the BIOS, if thats the case, once Nibitor supports it, there may be some headroom to do some Bios modifications.

Also I heard that the max core clock is driver invoked? if thats the case there could be a way to get around that as well but not easy, for example, if/when RivaTuner "officially" supports the card, providing you set rivatuner to startup in windows you can then diasable the NVdriver in "processes" in task manager, that may allow you to overclock the card further.......just a couple of what might be "fruitless" idea's

 

[trt740]

Unfortunatly, you cannot unlock what is laser cut, sadly I think the days of simply disabling hardware have well and truly gone. The next thing you need to look at is voltage, they may have undervolted the card, dump the bios and have a look at it in Nibitor, you should be able to read it at least but of course Nibitor won't let you write to the BIOS juust yet as there is no support for the GS yet.
Then google around for what the boards max voltage is....I am willing to guess that the boards can take/give a lot more than is regulated in the BIOS, if thats the case, once Nibitor supports it, there may be some headroom to do some Bios modifications.

Also I heard that the max core clock is driver invoked? if thats the case there could be a way to get around that as well but not easy, for example, if/when RivaTuner "officially" supports the card, providing you set rivatuner to startup in windows you can then diasable the NVdriver in "processes" in task manager, that may allow you to overclock the card further.......just a couple of what might be "fruitless" idea's

Already flashed it added 1.1v to extra made zero difference it is bios locked it will artifact as soon as you set it within ten MHz of 700Mhz core when you test it with ati tools. The new version of Nibitor will let you edit the bios. I have since flashed it back. I can bech alot higher, but not stable

 

[Tatty_Two]

Already flashed it added 1.1v to extra made zero difference it is bios locked it will artifact as soon as you set it within ten MHz of 700Mhz core when you test it with ati tools. The new version of Nibitor will let you edit the bios. I have since flashed it back. I can bech alot higher, but not stable

OK, disable the nv driver in taskmanger then and overclock the card higher in rivatuner and see if that does the job, if that dont work, ensure that rivatuner starts up with windows, then go into .....

Start > Run > type "msconfig" then go to the startup tab and temporarily disable the NV driver from starting up.

 

[Saakki]

this all nvidia shadyness gets actually pretty interesting ...thanks for wizz for sharing up knowledge.. finnish site reported this and its hot topic now

 

[trt740]

OK, disable the nv driver in taskmanger then and overclock the card higher in rivatuner and see if that does the job, if that dont work, ensure that rivatuner starts up with windows, then go into .....

Start > Run > type "msconfig" then go to the startup tab and temporarily disable the NV driver from starting up.

Will try this weekend have the Flu so I'm just barely hanging on

 

[Scyphe]

I apologize for not understanding the article then.

So if Riva Tuner reads the real clock value, which in my case, my core clock stock is 650Mhz, but shows as 729Mhz in Riva Tuner is the real clock speed of the card. Then this makes no sense because "ANY" clock speed over 650Mhz for the 9600 GT OC VOIDS THE WARRENTY..

Rivatuner does not show the real clock value since the 9600GT is NOT using the standard 27Mhz crystal to calculate it's frequency. However, and this is very important, the fact that it shows WRONG values makes you question WHY they are wrong since it's using a standard way of deciding frequencies. And here's the deal. nVidia have changed the way 9600GT sets it's frequency. Instead of using the crystal on the card it uses PCIe-bus/4. That's why the card becomes overclocked when you increase the PCIe-bus beyond 100Mhz. And this is why Rivatuner is a symptom of this change. The 9600GT use a multiplier of 26 with the PCIe-bus frequency/4. This means that at 100Mhz PCIe-bus the frequency of the core will be 26*25=650Mhz. It also means that if the PCIe-bus is increased to for instance 110Mhz the core will run @ 26*27.5(pcie/4)=715Mhz.

WHY nVidia have decided to change the way the 9600GT set it's frequency is the shady part. They haven't said a word, and when asked they've denied any knowledge. The result is that there's a lot of overclocked 9600GT's being reviewed as stock 9600GT's which will inflate the results and pretty much muddle any possible comparisons.

So what now? We have a card that sets it's core clock speed to that higher than the factory default. So if the 729Mhz core clock as shown by Riva Tuner, causes my card to burn up or malfuction, I am out $239.00 because Nvidia added some new feature or clock?

Pretty much except that Rivatuner isn't showing the true speed. To get the REAL speed you'd have to take 729, split it with 27, then multiply it with pcie/4. The fact that you paid $239 for a 9600GT instead of getting a 8800GT is something you'll have to live with.

So a Stock OC Card installed on a pc with the PCIe oc'd to 125Mhz automaticly voids my warrenty correct?

Depends on the warranty and the company policy of the vendor.

And no, overclocking the PCIe-bus doesn't overclock other cards, it ONLY overclocks the PCIe-bus.

 

[Steevo]

the basis of the article is that the majority of people who would potentially buy this card don't really mess around with OCing their systems. Seeing as how it's with a 9600 card, and I'd expect those to slide into the mid-range bracket 'ere long - the mid range market is "typically" the highest cost average consumers are willing to purchase on a new card.

Even still, there aren't too many users that really start digging into a system BIOS for graphics OCing, and remember, there are only a few brands of motherboards whose BIOS allows for adjusting the PCIE frequency. If you have a OE system from Dell, HP, eMachines, etc, you wouldn't have access to that setting in BIOS at all, either. And being such, if the setting is left to [AUTO] or there isn't a means to adjust it, the PCIE frequency can change during running of applications (someone correct me on that if I'm wrong).

What it all boils down to, is that for the average consumer purchasing a 9600 for use in the Dell, Gateway, etc setup - or users who don't OC and just run everything at stock speeds - your card will effectively OC itself without your knowing about it.

I'm personally not too keen on the idea, as it makes the card appear better than it truly is, and effectively throws scoring for reviews because the card is OCing itself. IMO, this is like steroid use amoungst athletes - until there's hard evidence that something is amiss, no one's the wiser about it, but that doesn't make it right.

180 degrees wrong. In most systems the card sucks for the $$$$$ spent compared to competing products. Most systems will leave the PCI-e at 100Mhz locked and the card will run at stock where it failes to meet the foretold performance.


Nvidia conpares it to a HD3870 in http://www.nordichardware.com/news,7357.html when they showed it to nordichardware and announced it would trounce the 3870. But when a first stock to stock comparison is made.
http://www.tweaktown.com/reviews/1293/4/page_4_test_system_setup_and_3dmark06/index.html it manages to edge ahead a small amount in lower res, and fall behind in higher res.


But when placed in a linkboost board that overclocks the core with out the user knowing, it would edge out the HD3870.


Again if we compare stock to stock http://www.techpowerup.com/reviews/Biostar/GeForce_9600_GT/20.html the HD3850 512 kicks its ass at stock, making it the better buy for high res gaming. But if we compare a overclocked version. http://www.techpowerup.com/reviews/VVIKOO/GeForce_9600_GT_Turbo/20.html The card suddenly looks better.



So what Nvidia has done is to show a card that when overclocked unknown to reviewers and owners has better performance than a competing card. But in most owners and reviewers boards failt to reach said performance. Now we have to wonder if they underclocked the card for the reason of core binning problems, issues with heat, or some other form of deception.


One thing remains however, this card is not what it was or is marketed to be in stock form.

 

[imperialreign]

180 degrees wrong. In most systems the card sucks for the $$$$$ spent compared to competing products. Most systems will leave the PCI-e at 100Mhz locked and the card will run at stock where it failes to meet the foretold performance.


Nvidia conpares it to a HD3870 in http://www.nordichardware.com/news,7357.html when they showed it to nordichardware and announced it would trounce the 3870. But when a first stock to stock comparison is made.
http://www.tweaktown.com/reviews/1293/4/page_4_test_system_setup_and_3dmark06/index.html it manages to edge ahead a small amount in lower res, and fall behind in higher res.


But when placed in a linkboost board that overclocks the core with out the user knowing, it would edge out the HD3870.


Again if we compare stock to stock http://www.techpowerup.com/reviews/Biostar/GeForce_9600_GT/20.html the HD3850 512 kicks its ass at stock, making it the better buy for high res gaming. But if we compare a overclocked version. http://www.techpowerup.com/reviews/VVIKOO/GeForce_9600_GT_Turbo/20.html The card suddenly looks better.



So what Nvidia has done is to show a card that when overclocked unknown to reviewers and owners has better performance than a competing card. But in most owners and reviewers boards failt to reach said performance. Now we have to wonder if they underclocked the card for the reason of core binning problems, issues with heat, or some other form of deception.


One thing remains however, this card is not what it was or is marketed to be in stock form.

I guess I must have misunderstood the report a bit . . . so, the 9600 auto OCing only happens on nVidia chipset based boards, and we wouldn't see that same increase in clocks on boards from other manufacturers with different chipsets that utilize automatic PCIE freq clocking (i.e. ASUS boards with PEG Link controls)?

Even still - it's a bit of an underhanded practice . . . it's too bad how fanatical the fanboish base has become, as a few years ago practices like this hurt the company for quite a long while

 

[cbunting]

On May 23rd, 2006, Nvidia released a press release about the linkboost technology.
http://www.nvidia.com/object/IO_31319.html

The Techpowerup article was written Feb 29th, 2008.

All in all, it's not a new feature and has been out for over a year and a half. I mean, what is the real point of the article? Simply that it took a review 1 1/2 years later for someone to notice?

You guys have not once understood any of the replies I've made since the very first one. Either that, or you all must have not known about any of these features either it seems. I don't see how you consider this, New, Shady or having anything to do with the 9600.
Chris

BTW: From what I can find online. Both the nForce 590i and Linkboost as mentioned in the article have been discontinued.

 

[Solaris17]

um....idu what that post has anyhting to do with this article.....linkboost isnt the issue being discussed here....of couse this article wouldnt come up till a year and ahalf later because it was at the release of the 9 series were the pci-e freq was used to determine core clock....

 

[cbunting]

it was at the release of the 9 series were the pci-e freq was used to determine core clock....

But that is the problem. It doesn't.

 

[Scyphe]

On May 23rd, 2006, Nvidia released a press release about the linkboost technology.
http://www.nvidia.com/object/IO_31319.html

The Techpowerup article was written Feb 29th, 2008.

All in all, it's not a new feature and has been out for over a year and a half. I mean, what is the real point of the article? Simply that it took a review 1 1/2 years later for someone to notice?

You guys have not once understood any of the replies I've made since the very first one. Either that, or you all must have not known about any of these features either it seems. I don't see how you consider this, New, Shady or having anything to do with the 9600.
Chris

BTW: From what I can find online. Both the nForce 590i and Linkboost as mentioned in the article have been discontinued.

Damn you're daft.

Ask yourself: every single one in here (especially w1zzard, developer and author of ATITool and Unwinder, developer and author of Rivatuner) EXCEPT you grasp the simple, explained (times 10) and proven issue with 9600GT calculating it's core frequencies differently from earlier generations. Somehow you just don't get it, you're not even on the same planet as the rest of us. Do us all a favor and sell your PC and buy a Wii. You clearly can't grasp even the simplest of concepts since you're spamming with all sorts of totally unrelated and useless links that have nothing to do with the issue we're discussing. OR you're a troll trying to defend the 9600GT like a knight (although in this case it's more like Don Quijote since you don't even know what you're defending it for or against, you're off somewhere else with us just looking in amazement while you fight imaginary monsters).

You guys have not once understood any of the replies I've made since the very first one.

You didn't even understand what we're discussing, hence your replies made absolutely no sense to anyone. It's like trying to explain quantum physics to a retard.

 

[Solaris17]

But that is the problem. It doesn't.

are you serious? i might know

<---------------------



and seeing as i got my 9600 b4 release youd think i talked to the programmers?

 

[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...

Changes in GPU-Z version 0.1.7:
Corrected several GeForce 9600 GT readings

Chris

 

[Solaris17]

RT supports it you have to add a line. hold on ill try some tests in a bit.

 

[cbunting]

It doesn't as I already posted..

From Alex, AKA Unwinder / Author of Riva Tuner

I'd like to add that v2.06 "knows" nothing about G94 core has no internal G94 specific codepath. So adding G94 support to 2.06 this way may cause unpredictable results and several thing may function improperly or not work at all. For example, RAM type will be detected improperly, bus width will not be detected at all, core clock can be monitored improperly in hardware monitoring module etc.
The only case when it is safe to add new card support by means of editing GPU database in .cfg file is when you're adding support for new display adapter model based on supported GPU family (e.g. you can safely add 8800GTS 512 support this way, because 2.06 fully supports G92 core). G94 is a bit different story. So please use it at your own risk.

http://i30.tinypic.com/ivd2mo.jpg

The article was based on results from software that doesn't support the card.

I can not reproduce the results. This is why I have asked all of those who have a 9600 to post theirs. We have one author who posts a review about this feature of the 9600. One person who has noticed this. And now I can't reproduce the result. So either more people get the same affect of the article for proof, or no one else can reproduce it either. For myself, I can't. It makes no difference what I OC as nothing changes the clock speeds.

 

[Solaris17]

well i guess im lucky it worked huh?