-
We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.
ATI A Cheater
Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
I hate stupid people too. But I'm just trying to find out the truth like many of you out there are too.
I owned an Radeon 8500. I thought it was a great card although some early drivers left some textures out in some games and trueform never really worked properly. I also owned a Geforce 4 ti 4600, and this card was great too. And now I own a Voodoo5 5500 and this card is great, and its drivers are heaven compared to others.
All of these companies are great. But I don't think it was right for everybody to jump on Nvidia, while noone questioned ATI's optimizations. I am even guilty of jumping on Nvidia, but the information was presented wrong to begin with and many ifluences of ignorance were out there. ATI was cheating all this time and we didn't even know, or did we care? I've even heard that ATI uses bilinear filtering in ut 2k3 instead of trilinear that Nvidia uses.
I'm not rich and I can't afford to make a mistake, I want what's best for my money. And right now, ATI is doing more cheating, and more hardware and driver issues are popping up. The 5900 is not the best performer but it will rock the games I would like to play right now and its cheaper than the 9700 Pro, who after a year is only 100 dollars cheaper in best case senario.
It's clear and it's cut, Nvidia works and ATI does not... at least for now.
I owned an Radeon 8500. I thought it was a great card although some early drivers left some textures out in some games and trueform never really worked properly. I also owned a Geforce 4 ti 4600, and this card was great too. And now I own a Voodoo5 5500 and this card is great, and its drivers are heaven compared to others.
All of these companies are great. But I don't think it was right for everybody to jump on Nvidia, while noone questioned ATI's optimizations. I am even guilty of jumping on Nvidia, but the information was presented wrong to begin with and many ifluences of ignorance were out there. ATI was cheating all this time and we didn't even know, or did we care? I've even heard that ATI uses bilinear filtering in ut 2k3 instead of trilinear that Nvidia uses.
I'm not rich and I can't afford to make a mistake, I want what's best for my money. And right now, ATI is doing more cheating, and more hardware and driver issues are popping up. The 5900 is not the best performer but it will rock the games I would like to play right now and its cheaper than the 9700 Pro, who after a year is only 100 dollars cheaper in best case senario.
It's clear and it's cut, Nvidia works and ATI does not... at least for now.
I think he is talking about the cheating claims made by him, of which one of them showed up on the NVnews forums before the reviews were even out and debunked by the poster particleman who showed the "cheat" does not affect benchmark scores
, and can be disabled through 2 variables in the ut2k3 startup .ini file
Corporate Thug
Lifer
i dont see any clear evidence of cheating
let me just post this link...
make up your own mind..
did lars take the shots himself or did he just post stuff he was given... or... does he have an inherent inability to turn the brightness up and check again to see what is being rendered and what is not being rendered ?
check this out
work was done by chris W @ www.nvnews.net... just turn up brightness and lo and behold... the "oh its not rendering therefore its cheating" statement looks a little daft 🙂
like I was saying about the big 3 and what their obligations are to consumers who look to their sites for information...
make up your own mind..
did lars take the shots himself or did he just post stuff he was given... or... does he have an inherent inability to turn the brightness up and check again to see what is being rendered and what is not being rendered ?
check this out
work was done by chris W @ www.nvnews.net... just turn up brightness and lo and behold... the "oh its not rendering therefore its cheating" statement looks a little daft 🙂
like I was saying about the big 3 and what their obligations are to consumers who look to their sites for information...
DaveBaumann
Member
If the UT2003 filtering being talked about is just the difference between their control panel and application filtering, this is hardly news - we've talked about this months back and have been benchmarking with application controlled AF (including altering the .ini file in UT2003) where ever we can:
http://www.beyond3d.com/reviews/ati/9800_256/index.php?p=13
As for Aquamark, I'm not sure that showing an image being rendered slightly different on one architecture (with an internal precision of at least FP24) to an image rendered on a completely different architecture (with an internal precision of 16) really tells you much. The starting point is the reference rasteriser and then you see which is closest.
http://www.beyond3d.com/reviews/ati/9800_256/index.php?p=13
As for Aquamark, I'm not sure that showing an image being rendered slightly different on one architecture (with an internal precision of at least FP24) to an image rendered on a completely different architecture (with an internal precision of 16) really tells you much. The starting point is the reference rasteriser and then you see which is closest.
DaveBaumann
Member
I'm not sure how to deal with that! 😕
Jeff7181
Lifer
I'm sick of hearing all this cheating stuff... who says it's cheating?
Are we agreed that the meaning of cheating is to break the rules to gain an unfair advantage?
Assuming we all agree on that... if both nVidia and ATI are using methods of reducing image quality to gain performance, is it really cheating? Are they gaining an unfair advantage if they both do it? Or is it just competition and different methods of achieving similar results?
I don't think it's cheating until one of them somehow figures a way to say, multiply the 3DMark score by 1.2 behind the scenes or something like that. Reducing the image quality of certain things to gain performance, and have little or no effect on what you can see is smart in my opinion. And not rendering something that isn't even displayed on the screen is also smart in my opinion.
Are we agreed that the meaning of cheating is to break the rules to gain an unfair advantage?
Assuming we all agree on that... if both nVidia and ATI are using methods of reducing image quality to gain performance, is it really cheating? Are they gaining an unfair advantage if they both do it? Or is it just competition and different methods of achieving similar results?
I don't think it's cheating until one of them somehow figures a way to say, multiply the 3DMark score by 1.2 behind the scenes or something like that. Reducing the image quality of certain things to gain performance, and have little or no effect on what you can see is smart in my opinion. And not rendering something that isn't even displayed on the screen is also smart in my opinion.
jiffylube1024
Diamond Member
I don't know what to say. I looked at most of the pictures and couldn't really find too many differences, although a few were present.
On that one picture with the fog, yes it looked like there was less fog, but nothing to directly indicate blatant "cheating". The fog was still there, it still looked perfectly fine. It's not like they actually entirely removed the fog for one scene, or clipped the other planes.
On the Aquamark Anisotropic page, the second picture looks a lot nicer on the Nvidia card. Is ATI cheating, or is it a problem with Aquamark? Maybe it's a cheat, I have no idea. Everything else is identical.
Pixel Shader ATI "overdraw bug" . There's no proof, just Tom's word. Good thing he has a lot of credibility, eh
.
Check out this page of Tom's "review" where he says ATI's AA in this picture looks substantially better than Nvidia's. Look at Nvidia's AA. It looks like crap. As a matter of fact, it looks like Nvidia is using lower quality AA. Couldn't this itself be considered "cheating"? Is THG just skewing everything like they always do and only analyzing the data that fits their theory?
If Tom actually had any credibility, he'd be treading on eggshells for some of his claims. Fortunately for him, he has none.
On that one picture with the fog, yes it looked like there was less fog, but nothing to directly indicate blatant "cheating". The fog was still there, it still looked perfectly fine. It's not like they actually entirely removed the fog for one scene, or clipped the other planes.
On the Aquamark Anisotropic page, the second picture looks a lot nicer on the Nvidia card. Is ATI cheating, or is it a problem with Aquamark? Maybe it's a cheat, I have no idea. Everything else is identical.
Pixel Shader ATI "overdraw bug" . There's no proof, just Tom's word. Good thing he has a lot of credibility, eh
Check out this page of Tom's "review" where he says ATI's AA in this picture looks substantially better than Nvidia's. Look at Nvidia's AA. It looks like crap. As a matter of fact, it looks like Nvidia is using lower quality AA. Couldn't this itself be considered "cheating"? Is THG just skewing everything like they always do and only analyzing the data that fits their theory?
If Tom actually had any credibility, he'd be treading on eggshells for some of his claims. Fortunately for him, he has none.
you confuse cheating/optimizations in real world games with the same in benchmarks...
benchmarks are designed to stress certain things... not rendering something in that test is going to give false results...
in games... do whatever... occulsion culling already occurs for both ati and nvidia when it comes to rendering objects in real time...
gaining an unfair advantage in BENCHMARKS is incorrect... doing global improvements in games on the other hand is a different matter 🙂
Originally posted by: jiffylube1024
I don't know what to say. I looked at most of the pictures and couldn't really find too many differences, although a few were present.
On that one picture with the fog, yes it looked like there was less fog, but nothing to directly indicate blatant "cheating". The fog was still there, it still looked perfectly fine. It's not like they actually entirely removed the fog for one scene, or clipped the other planes.
On the Aquamark Anisotropic page, the second picture looks a lot nicer on the Nvidia card. Is ATI cheating, or is it a problem with Aquamark? Maybe it's a cheat, I have no idea. Everything else is identical.
Pixel Shader ATI "overdraw bug" . There's no proof, just Tom's word. Good thing he has a lot of credibility, eh.
Check out this page of Tom's "review" where he says ATI's AA in this picture looks substantially better than Nvidia's. Look at Nvidia's AA. It looks like crap. As a matter of fact, it looks like Nvidia is using lower quality AA. Couldn't this itself be considered "cheating"? Is THG just skewing everything like they always do and only analyzing the data that fits their theory?
If Tom actually had any credibility, he'd be treading on eggshells for some of his claims. Fortunately for him, he has none.
they have the raster images posted @ b3d...
ati and nvidia in general are rendering per specification... both are a lot darker but apparently this has something to do with ingame gamma or something.. I didnt read that much in it... I was checking out the imgs...
slight degradation with some objects on nvidia's shots with 51.75... no 52.xx shots to compare with... ati rendered same as raster..
DaveBaumann
Member
Actually, just using a lower quality mip level on the particles for would result in the fog looking thicker.
jiffylube1024
Diamond Member
^ Wow. I was looking at the fog, and in the pics @ Tomshardware, there is no question that the ATI does look slightly less dense, however the area of fog is identical (pay particular attention to where it matches up on the wall vs. the Nvidia card. It's the same).
So, I did the same thing that was in that link essentially to prove it to mysefl: I opened up the PNG's in Photoshop, copied it and pasted the original on top and the one I would edit on the bottom. I then upped the brightness on only ATI's image (you can see the whole image is brighter: the jeep, the walls, etc to show I'm not just copying Nvidia's) and the fog now looks almost identical. Indeed, the only problem with ATI's rendering is that the brightness is a bit too low, this is definately not a cheat!!!
Edit: oh yeah the pic is here where I upped the brightness. Check it out.
Is there any reason why Tom's "uncompressed PNG" files linked are the exact same thing as the ones in the article? I guess they're just being lazy...
Well, I'm surprised noone has bothered to mention the fact that the 52.16's are missing a lot of DX9 functions.
Most importantly... No Phong/Ansitropic lighting, no Phong spot lighting, no gaussian or cross blurs, no edge detection shader or water colour shaders.
Admittedly, most games won't use half the visual improvements (like water colour shader might only show up when there is blood in the water) But Phong/AF lighting and spot lighting is used quite a bit in DX9 games. People are going to start noticing that their Nvidia card and brand new 52.16 drivers are missing a lot of these "basic" DX9 features.
It would be sad if Nvidia users had to wait another six months just to get up to speed on basic DX9 Phong lighting (and possibly take a performance hit turning it on too)
Most importantly... No Phong/Ansitropic lighting, no Phong spot lighting, no gaussian or cross blurs, no edge detection shader or water colour shaders.
Admittedly, most games won't use half the visual improvements (like water colour shader might only show up when there is blood in the water) But Phong/AF lighting and spot lighting is used quite a bit in DX9 games. People are going to start noticing that their Nvidia card and brand new 52.16 drivers are missing a lot of these "basic" DX9 features.
It would be sad if Nvidia users had to wait another six months just to get up to speed on basic DX9 Phong lighting (and possibly take a performance hit turning it on too)
hominid skull
Senior member
Johnbear007
Diamond Member
Using Tom's as evidence is like using the enquirer or daily globe 😛
Actually i'd believe in the Yeti or Aliens from those rags before I believed any of the pure crap spewing out of Tom's
Johnbear007
Diamond Member
Both companies cheat, period. anyone who cant see that is just another tard fanboy
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 25K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 24K
-
-
