Originally Posted by lord_shar

I have a 6800gt and 6800Ultra, both from BFG. If given a choice between the X800XL vs. the 6800GT, I'd probably go with the nVidia for several reasons:

1) the 6800gt supports full DX9.0c, while the ATI X800 only goes up to DX9.0b. This doesn't matter with today's games, but DX9.0c might become a requirement for some future titles.

2) ATI's drivers tend to perform dynamic texture quality adjustments during high stress video movement, lowering image quality in order to maintain fps. While this is an intelligent way of managing hardware resources, it doesn't allow accurate measurement of the hardware's performance until the feature is turned off. When it is, more than a few titles bench lower on ATI boards vs. nVidia.

However, the nVidia 6800 series, especially the Ultra, are huge power hogs. If you don't have a sufficiently beefy power supply, you may encounter some system stability issues if you go with one of these power-hungry video cards.

Originally Posted by Zakarr

1) DirectX 9.0b and 9.0c are mainly bug fixes and don't define any new 3D hardware feature standards so both ATI and NVIDIA support DirectX 9.0 but X800 series support pixel shader 2.0b and vertex shader 2.0 while NVIDIA GF 6800 series support pixel and vertex shaders 3.0

Higher version shaders are useless if game don't support them. You won't automatically gain any image quality with better shaders.

Originally Posted by Zakarr

2) You can turn it off if you want.

Originally Posted by Zakarr

Benchmarks are no the whole truth. Results don't represent all the games in the world. Some games work better with ATI while some other games work better with NVIDIA. One might to want his/her favorite games work the best way rather than blindly looking for benchmark results from few games.

Originally Posted by Zakarr

ATI consume less power which means less heat and less stress for power supply. Usually ATI cards have more silent stock coolers at least in certain brands and you can usually reduce the fan speed without overheating a lot more than in NVIDIA GF 6800 to reduce cooler noise.

ATI X800 has better quality anti-aliasing than in NVIDIA GF 6800. And better anti-aliasing performance with high resolutions like 1600x1200 or maybe even with 1280x960 or 1280x1024

NVIDIA usually have better OpenGL performance so Linux users and OpenGL graphics based games like Call of Duty, Doom3, Quake series, Linux 3D games and all games which are based on Quake or Doom 3 engine, are faster and/or less buggy.

NVIDIA often overclocks better so overclocker might want to choose it instead ATI. They are most likely more future proof cards because of more advanced shaders but no one know how soon and widely these shaders are going to be used in near future games and how much gamers get benefits from them. Latest Splinter Cell and Far Cry support these shaders but in Far Cry the cap between X800 and GF6800 is pretty minor. However, latest Splinter Cell gains noticeable image quality differences with GF 6800 but that is because from some reasons developers didn't add support for ATIs 2.0 shaders and ATI have to use old 1.1 shaders. Maybe it was done in purpose by offering bribes. Who knows.

Originally Posted by MasterQu

Couldn't catch me using ATI. Nvidia sold me when I had to call them for driver support. As I was talking to them they rebuilt the driver, I was using on a (i think) 5700 card at the time that was having issues with UT2004, and emailed it to me before i hung up the phone. Thats service and it bought my loyalty. Never mind the fact that everytime ati comes out with some thing nvidia reciprocates a pwns ati a$$.

Originally Posted by lord_shar

Yep, here's the link from Tom's Hardware Guide:

http://graphics.tomshardware.com/gra...603/index.html

The article from June 2004 found that the X800's catalyst drivers dropped the texture filtering mode from trillinear to bi-linear mode during high load situations, but ramped up to trillinear mode once GPU load went down. Although ATI insists that the X800 is still performing tri-linear filtering, it's actually "bri-linear" since it's mixing the two filter modes. This filtering could not disabled until many several driver revisions later. As far as I know, it is still present in the current ATI catalyst driver set.

Although the technique does have its merrits, it is completely software driven and doesn't paint an accurate portrait of the hardware's normal performance when benched against a "non-driver-optimized" video card.

Originally Posted by Dirkiess

Algren, I read something just recently regarding the Crossfire series. I'll have to see if I can find the article, but it was related to the same thing, but it also mentioned that the drivers were only in Beta and not the final release.

Whether this was true or not, I don't know, but it's either plausible or a possible side step for ATI until they can get the drivers up to speed.

To be honest though, I think there may well be pro's and Con's for both ATI and NVidia.

These could all be related to the systems that people have and a prime example of this, was a news article I read today that says NVidia's own NForce 3 motherboards cause problems with there own 6800 series of cards. Now how is that for compatibility?

I'll see if I can find either of the links to the above articles. I've read a lot of articles today.

Originally Posted by MaglorD

From what I read in the article, both manufacturers heavily optimise their drivers to get the most out of the hardware. This is to be expected since delivering true filtering slows down the hardware considerably. The trick is not to sacrifice image quality too much while employing the optimisations and this ATI appears to have accomplished. Don't forget how horrible Nvidia's FX optimisations were.

Most review sites do not disable Nvidia's optimisations while leaving ATI's optimisations intact, so I don't see how the comparisons are unfair.

It is generally accepted the X800XL is comparable to the 6800GT in fair comparison, without disabling optimisations while running either card. So pick whichever is cheaper.

Originally Posted by lord_shar

The reason nVidia's 6800-series didn't get slammed was because they maintained full anisotropic filtering and consistent image quality throughout the tests . ATI's x800's, on the other hand, dropped image quality for the sake of frame rates without informing its user-base until complaints came in. I'm just glad ATI responded positively to the criticism by adding an off-switch to disable bri-linear filtering. Unfortunately, this also revealed ATI's X800 hardware design starting to show its age vs. the 6800 series. I just hope they do better against the 7800GTX with their next solution sans "cheating."

Originally Posted by MaglorD

I truly doubt any card performs trilinear filtering even with bri-linear turned off. So it's pointless to compare without optimisations. The important thing is to ensure optimisations do not degrade quality compared to the competition.

Originally Posted by MaglorD

The reason Nvidia didn't get slammed is because ATI went and told everyone they were performing true trilinear filtering when they actually optimised by using brilinear. Even so, the image quality with ATI's brilinear filtering is superior to Nvidia's brilinear filtering and does approach true trilinear filtering.

Originally Posted by Techie

To be honest its any mans game. ATI and Nvidia have their ups and downs. Once Crossfire is used to its full potential I bet it will blow SLI away. ATI is known for being late to release cards, but they never disappoint.

Originally Posted by lord_shar

Speculation is fine, but please provide some links backing up the above to at least support your statements and reasoning.

I have yet to read any articles mentioning any bri-linear filtering being done by NVidia. Can you please link the URL here if you have it?

Yes, I'm sure Nvidia does have some degree of video driver optimization. However, we users cannot detect these optimizations since they don't leave any visible image degredation, while their ATI counterparts do.

Originally Posted by MaglorD

Quoted from the Tom's Hardware article.

This is indeed a "cheat" that both major vendors now do. Instead of always sampling the two adjacent mip map levels and doing a full blend between them, they have plateaus where only a single mip level is sampled, reducing the average samples from 8 to about 6. It is actually a pretty sensible performance enhancement, with minimal visual issues.

I'm not sure what visible image degradation you are referring to.

Here is what is quoted in the Tom's Hardware article:

The objective of trilinear filtering is to make transitions between mipmap levels as near to invisible as possible. As long as this is achieved, there is no "right" or "wrong" way to implement the filtering.

We have added intelligence to our filtering algorithm to increase performance without affecting image quality. As some people have discovered, it is possible to show differences between our filtering implementations for the RADEON 9800XT and RADEON X800. However, these differences can only be seen by subtracting before and after screenshots and amplifying the result. No-one has claimed that the differences make one implementation "better" than another.