Nvidia Driver Cheats Editorial
 

Hardware Analysis has their take on Nvidia's driver "optimizations". Its unfortunate that every time Nvidia releases a new product this occurs. Good product, bad product, or in the middle, it presents a doubt over the quality and actual performance of the product.

I'm sure you've read reports around the web about Nvidia allegedly cheating with their yet to be released 60.72 build drivers. These drivers were sent out to journalists evaluating the new GeForce 6800 Ultra and the NV4x architecture. Just two short weeks after the first reviews were posted one website published a lengthy article discussing image quality and how some frames rendered by the new GeForce showed differences from the same frames rendered on ATi hardware. So is this the proverbial smoking gun we're looking at, is Nvidia indeed making shortcuts to boost the performance of their new architecture or is something else going on here?

heh....funny

sad that they have to cheat...

Very true,but if you read the article can you really call it cheating. Seems that ATI and Nvidia do the same thing, just different takes on how to render an image.

Jury's still out.

Yeah.. I remember ATI got slammed too along with nVidia back when 3dmark2k3 was released for "optimizations".

except nvidia does it over and over again.. ati only got caught once

Meh.. Im just leaning back here and watching them duke it own.. Whoever wins the benchmarks wars for PCI-Express, I'm just gonna choose that card.. I really don't care if its ATI, nVidia, s3.. Not really a fanboy when it comes to GPU's.

oh really...


anyway, s3 has no chance and should go die, nvidia should play by the rules and ati should stop being snotty...

First off I like rooting for the underdog, sometimes they end up making the superior product (AMD).

Secondly the only reason people think Nvidia doesnt play by the rules is because others say so, If you think ATI isisnt or hasnt tried the same thing, then your wrong.

Thirdly ATI has every reason to be snooty, they make the best product on the planet, for a lower cost than most, while producing better image quality than most.

The jury is out on IF and HOW much better the 6800Ultra will be and if they really are cheating. Unless your a vid guru I tend to try educating myself before I hack into a product.

My personal Opinion: S3 has yet to produce anything worth buying that was competitive in performance. Nvidia should stop trying to cheat the benchmarks and their consumers, and ATI is starting to act holier than thou.

But those are just my opinions. :)

ATI is holier than thou...They are superior. I've quit Nvidia after going through 2 cards in 4 months, I've had my card for over a year now and I'm still happy.

i would like tpo point out that the underdog is better motivated to win, so well see . ..

yea if I was nvidia i'd be pretty pissed off.. and as far as I can tell they are (and I like that in a card maker.. until they get unethical about it)

nvidia is doing fine... besides... they were raking in the dough back then from the ti gf 4 series, gf2, gf3, and even the gf 1..... i have a feeling their next line of cards will prosper nicely...

The only major allegation against nVidia was for the last generation of cards, the first FX's. As Jason said, it turned out that ATi had been doing the same exact thing, and yet they didn't get such a bad rap and aren't the ones thought of immediately when the topic comes up because they weren't the ones caught first.

Now, in my opinion, this "cheating" isn't at all a bad thing on the part of either company. The drivers are developed so that the benchmarks use the hardware in their most efficient manner. The reason I don't feel this is bad, and actually a good thing, is so that we can see how the hardware itself performs, and not how it performs under a specific version of some software. When DX gets optimized with a 9.0b or c , perhaps, you will already know which card is faster because the unoptimized software isn't the bottleneck in those results, the hardware is. And isn't that what we are trying to compare after all?

i have a tendency to agree with vee_ess. it's better to test the maximum level rather than some random range in the upper middle. on the other hand, if corners are being cut so that the hardware is just not performing the test, then that's not good. It's like a runner running half the race and saying he won b/c he stopped running first. well, he didn't run the whole race.