Notices

Overclockers Forums > Hardware > Video Cards > nVidia GPUs
nVidia GPUs
Forum Jump

Nvidia Caught Cheating AGAIN?!?!?

Post Reply New Thread Subscribe Search this Thread
 
 
Thread Tools
Old 08-31-05, 07:39 PM Thread Starter   #1
LoneWolf121188
Member

 
LoneWolf121188's Avatar 

Join Date: Dec 2004
Location: Sheppard Air Force Base, TX

 
Nvidia Caught Cheating AGAIN?!?!?


Quote:
Originally Posted by The Inquirer
WE CALL IT optimisation these days, but what do you call it when a firm over optimises? We used to call it something different in the days of the NV30, when Nvidia had to do something to make its chip work better and do better in benchmarks but to our surprise it seems to have done it again. The chaps from 3D Center, a very talented in-depth site, spotted and tested and proved that Nvidia is using a lower anisotropic filtering quality than any other card available.
Those guys noticed a texture shimmering problem when you are using normal driver settings. This was the case with NV40 cards but you could resolve this flickering by using high quality driver settings. This won't work on G70 based cards, so the guys well known for its thorough benchmarks went digging a little deeper into the chip.

It turns out that Nvidia is not doing anisotropic filtering the way it should and that the picture quality is the one to suffer. You will get the shimmering effect on your textures whenever you are using Geforce 7800GTX cards but you won't see this using Radeon cards.

The guys claimed that all NV40 and G70 cards will suffer from the same flickering problem and that these cards have "by far worse AF quality". They also add that Nvidia got the flickering because it was using general under sampling and as a result is getting the flickers. It's interesting to note that older Geforce 5800 Ultra won't suffer from this, just the new cards that 6800 or 7800 based.

Another German web site Computerbase , went a step further. It made a custom driver by changing the inf, where the driver could not recognise 7800GTX and use its optimisations. The card was listed as unknown but was working just fine. But when the guys went testing they noticed a massive performance drop when using those drivers, close to 30 percent and related it to anisotropic filtering. Nvidia has a lot to explain.

3DCenter, for the original article, is here in English while the Computerbase German article is here. We will ask Nvidia what is going on but we think there's something up. At lease the guys proved it isn’t a hardware bug - it's a driver problem only but performance drops dramatically as soon as you resolve it. µ
http://www.theinquirer.net/?article=25807

nvidia...c'mon guys, don't do this AGAIN...

What's your take?

Theres a follow up article here.

__________________
mATX Desktop (Build Log Here): Core i7 4770K @ 4.5 GHz, 1.3V, Maximus VI Gene, 16GB DDR3-1600 RipjawX, EVGA GTX780, Intel 520 240GB, Seagate 2TB, EK CPU and GPU waterblocks, MCP35X, 120.1 and 180.1 radiators.
Microsoft Surface Pro 3.

The Hardware Destruction Thread|OCForums Musicians Club Member -12|OCForums Metal Club Member -12

Well, if he opens the box and its dead, then it would be his fault because the state of the cat is determined by the observer. Then we could be all ‘n00b, you collapsed the wavefunction now! - RoadWarrior, on Schrödinger’s Cat
LoneWolf121188 is offline SETI Profile Heatware Profile   QUOTE Thanks
Old 08-31-05, 07:44 PM   #2
xTrEmEoVrClOcKr
member



Join Date: May 2004
Location: San Diego, CA

 
......

__________________
Core i5 2500K @ 4.6GHz w/ XIGMATEK SD1283 120mm
Asus P8Z68-Z
8GB Corsair Vengeance
640GB x 2 WD Black in RAID0, 500GB Data drive
Sapphire HD5830 @ 900Mhz / 1225Mhz
Lian Li V2000A w/ 550W Antec TruePower II
xTrEmEoVrClOcKr is offline   QUOTE Thanks
Old 08-31-05, 07:46 PM   #3
johan851
Insatiably Malcontent
Senior Member

 
johan851's Avatar 

Join Date: Jul 2002
Location: Seattle, WA

10 Year Badge
 
Quote:
NVIDIA WAS very fast to react to a story we pixellated yesterday. We wrote here about Nvidia's over optimisation. Well, as expected Nvidia said it's a bug and it could not be fixed by changing to high quality driver setting, which was the case with Geforce 6 series cards
So nVidia already has a new beta out to fix it. That was fast. Strangely fast...

The FX series should prove that nVidia knows how to implement AF, and the new driver proves that they knew exactly what to do to fix it. They must have, to fix it so quickly. From that evidence I think it's pretty clear that this was no accident.

Pathetic. Optimizations are one thing, but at the cost of image quality...oh well. It's nVidia. I'm not surprised.

__________________
ASRock Z68 Extreme3 Gen3 | 2500K @ 4.6GHz | 2x4GB Samsung DDR3 | GTX 750 Ti 2GB
120GB Crucial M4 | 2x 2TB Samsung F4 | Seasonic S12 600w
y2 DAC --> Custom M^3 --> Custom LM3875 ChipAmp --> Modula MTs
Dual Dell 2007WFP | Watercooled
johan851 is offline   QUOTE Thanks
Old 08-31-05, 07:57 PM Thread Starter   #4
LoneWolf121188
Member

 
LoneWolf121188's Avatar 

Join Date: Dec 2004
Location: Sheppard Air Force Base, TX

 
Quote:
Originally Posted by johan851
Pathetic. Optimizations are one thing, but at the cost of image quality...oh well. It's nVidia. I'm not surprised.
Agreed, though not quite with that last bit. However, if they keep pulling stuff like this...I've always been an nvidia fan, but the canadians up north are starting to look tempting...

__________________
mATX Desktop (Build Log Here): Core i7 4770K @ 4.5 GHz, 1.3V, Maximus VI Gene, 16GB DDR3-1600 RipjawX, EVGA GTX780, Intel 520 240GB, Seagate 2TB, EK CPU and GPU waterblocks, MCP35X, 120.1 and 180.1 radiators.
Microsoft Surface Pro 3.

The Hardware Destruction Thread|OCForums Musicians Club Member -12|OCForums Metal Club Member -12

Well, if he opens the box and its dead, then it would be his fault because the state of the cat is determined by the observer. Then we could be all ‘n00b, you collapsed the wavefunction now! - RoadWarrior, on Schrödinger’s Cat
LoneWolf121188 is offline SETI Profile Heatware Profile   QUOTE Thanks
Old 08-31-05, 08:00 PM   #5
xTrEmEoVrClOcKr
member



Join Date: May 2004
Location: San Diego, CA

 
Yeah, for sure .. I just got my 7800GTX and I hope the new betas dont drop my performance 30% .

__________________
Core i5 2500K @ 4.6GHz w/ XIGMATEK SD1283 120mm
Asus P8Z68-Z
8GB Corsair Vengeance
640GB x 2 WD Black in RAID0, 500GB Data drive
Sapphire HD5830 @ 900Mhz / 1225Mhz
Lian Li V2000A w/ 550W Antec TruePower II
xTrEmEoVrClOcKr is offline   QUOTE Thanks
Old 08-31-05, 08:09 PM   #6
Nandro
Disabled

 
Nandro's Avatar 

Join Date: Mar 2004
Location: Wherever I happen to be.

 
From those same rumor places the betas will fix it with nothing close to that kind of loss and only when in Quality mode.
Nandro is offline   QUOTE Thanks
Old 08-31-05, 08:12 PM   #7
techun
Member

 
techun's Avatar 

Join Date: Jul 2004

 
if people didnt care, then Nvidia just made their cards faster

stop nitpicking

__________________
Sugo SFF cube case
4850 45W
4 gigs ddr2 800
7200.11 500gig
techun is offline   QUOTE Thanks
Old 08-31-05, 08:19 PM   #8
Zulu-1
Member

 
Zulu-1's Avatar 

Join Date: Jan 2005
Location: Brampton, Ontario

 
do people have nothing better to do with their lives!?!
notice how no one else noticed beforehand... OR CARED?
why does it matter people...
(dont call me a fanboy...i have an ATI card, i just go with whats good/cheap)

__________________
THE BEAST: :mad:
Athlon64 2800+
Zalman CNPS7000B-AlCu w/ AS5
Chaintech VNF3-250 ~Nforce3 250~
Connect3D Radeon X800 Pro
1GB OCZ Premier
ThermalTake 420W PurePower
I am 49% addicted to Counterstrike.
Heat 2-0-0

CanadaLAN 2.0?
EA [Teamkilled] PC Gaming
Zulu-1 is offline   QUOTE Thanks
Old 08-31-05, 11:47 PM   #9
Meathead



Join Date: Jul 2004
Location: westminster, CO

 
I think people care because the 7800gtx costs $500+ and they spent that much so they could have a fast card with full quality.
Meathead is offline   QUOTE Thanks
Old 09-01-05, 06:00 AM   #10
techun
Member

 
techun's Avatar 

Join Date: Jul 2004

 
but if they were very happy before, what difference does this "cheap af" make?
if it works that well i'd want it on any card i buy

__________________
Sugo SFF cube case
4850 45W
4 gigs ddr2 800
7200.11 500gig
techun is offline   QUOTE Thanks
Old 09-01-05, 06:07 AM   #11
johan851
Insatiably Malcontent
Senior Member

 
johan851's Avatar 

Join Date: Jul 2002
Location: Seattle, WA

10 Year Badge
 
Quote:
but if they were very happy before, what difference does this "cheap af" make?
if it works that well i'd want it on any card i buy
It's the principle of it. If ATi's cards are rendering scenes at full quality, and nVidia's cards can't (or rather, the drivers aren't) render with the same quality, then every benchmark comparing the two is inherently flawed and unfair to ATi.

__________________
ASRock Z68 Extreme3 Gen3 | 2500K @ 4.6GHz | 2x4GB Samsung DDR3 | GTX 750 Ti 2GB
120GB Crucial M4 | 2x 2TB Samsung F4 | Seasonic S12 600w
y2 DAC --> Custom M^3 --> Custom LM3875 ChipAmp --> Modula MTs
Dual Dell 2007WFP | Watercooled
johan851 is offline   QUOTE Thanks

Post Reply New Thread Subscribe


Overclockers Forums > Hardware > Video Cards > nVidia GPUs
nVidia GPUs
Forum Jump

Thread Tools Search this Thread
Search this Thread:

Advanced Search


Mobile Skin
All times are GMT -5. The time now is 11:30 PM.
Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2014, vBulletin Solutions, Inc.
You can add these icons by updating your profile information to include your Heatware ID, Benching Profile ID or your Folding/SETI profile ID. Edit your profile!
X

Welcome to Overclockers.com

Create your username to jump into the discussion!

New members like you have made this the best community on the Internet since 1998!


(4 digit year)

Why Join Us?

  • Share experience
  • Max out your hardware
  • Best forum members anywhere
  • Customized forum experience

Already a member?