**thread name change* Nvidia and AMD moral and immoral business practices

[Lonbjerg]

I'd like to see this list. Part of the reasoning behind some of the posts in this thread is precisely that Nvidia has immoral business practices while AMD does not. If you can show that AMD has similar practices (i.e., that they are immoral in much the same way as Nvidia are), then those who preferred AMD products would have to realize that the grass is not greener on their side of the field.

You are the thrid poster I post this too directly, perhaps third time is the charm?

FarCry...AMD64 patch...look it up.

 

[blastingcap]

You are the thrid poster I post this too directly, perhaps third time is the charm?

FarCry...AMD64 patch...look it up.

Was that AMD or ATI who did that? That was a rhetorical q.

Anyway, if that's the worst that the combined AMD/ATI has done, then compared to NV, ATI is an angel and AMD is a saint.

 

[Scali]

I'd like to see this list. Part of the reasoning behind some of the posts in this thread is precisely that Nvidia has immoral business practices while AMD does not. If you can show that AMD has similar practices (i.e., that they are immoral in much the same way as Nvidia are), then those who preferred AMD products would have to realize that the grass is not greener on their side of the field.

Well, for starters, ATi was the first vendor to get caught at cheating in benchmarks (the Quack.exe incident).
AMD is still 'half-cheating' in benchmarks because their AI feature still reduces texture quality as a trade-off for speed. AMD advises reviewers to test with AI enabled, but they do not specify what AI does exactly. Hence most reviewers aren't aware of the reduction of image quality and benchmark with AI enabled.

And should we get into AMD's OpenCL/physics practices? For years now, AMD has been making promises about supporting OpenCL, and delivering an open physics solution. So far, nothing has happened. They don't even include an OpenCL runtime in the end-user drivers.
They've rigged some Havok/OpenCL physics tests, showing 'working' physics acceleration, but somehow even years later, nothing has emerged (the earliest set up was a X1900+X1600 system... that was AGES ago. Where is it?).
So basically they've only deluded their customers, in an attempt to keep people away from nVidia (who DO have OpenCL and physics for everyone).
(If you thought nVidia showing a mock-up card at the Fermi introduction was bad... what do you think of this then?)

There's plenty more, but you get the idea. Bottom line is, AMD/ATi aren't exactly 'sweethearts' themselves.

 

[Scali]

I could be wrong, but the way I interpreted that is this: you are accusing AMD of not wanting to do anything, just getting a free ride, and refusing NV's offer re: PhysX, yet you then acknowledge that by accepting such an offer AMD would be at the mercy of NV. So why then, would AMD ever accept such an offer from NV?

I already answered that: I think it's better for AMD to support PhysX than to not support any GPU-accelerated physics at all.

If they accept any such type of offer they might as well side with Intel since Intel's lead on AMD is larger than NV's lead on ATI--i.e., AMD/ATI has a better chance of dethroning NV than dethroning Intel. Under the circumstances I think it's pretty clear why AMD rejected NV, yet your first post obfuscated this and made it sound like ATI was being unreasonable, ungrateful, and lazy in refusing NV's offer.

I think AMD made the wrong choice. Why? Simple... What interest does Intel have in GPU-accelerated physics at all? None, right? It would only compete with their CPUs, that's not what they want. Intel would never actually release a Havok with GPU-acceleration, unless they would be certain that it runs better on their own hardware. Which is why nothing ever amounted from AMD's 'cooperation' with Intel/Havok on OpenCL. And that is probably also why AMD suddenly shifted focus from Havok to Bullet.

If they went with nVidia, at least they'd HAVE GPU physics. Perhaps not as good as nVidia's, but they may have a fighting chance by introducing faster hardware.

But AMD refused nVidia's offer... that is one part of the story.
Then AMD started attacking nVidia's PhysX for being proprietary, and tried to push OpenCL. Now, that is the other part of the story. If AMD says "You should port PhysX to OpenCL", that basically means: "You do the work for us" (especially since AMD always had the chance to work with nVidia and build an OpenCL implementation for PhysX themselves).

You then went on about how AMD passed up opportunities, but that is a) second-guessing AMD with the benefit of hindsight, even if you're right (and to some extent I think you ARE right) and b) irrelevant to the issue I had with your original statement which made it sound to me like you were saying ATI was being unreasonable, ungrateful, and lazy in refusing NV's offer.

I do think AMD was being lazy in not wanting to build the OpenCL implementation for PhysX themselves.

As for hindsight? I think you need to dive into the history a bit more. ATi was working with GPU-accelerated physics back in the days of the Radeon X1900. Basically ATi was more actively involved in physics/GPGPU than nVidia was, prior to nVidia's introduction of the 8800 series and the Cuda platform. So yes, I do think ATi missed the opportunity to acquire Ageia or Havok. They were there first, then they let both their competitors steal their thunder, and now AMD is the only one standing there empty-handed.

By the way, you'd be surprised at how much technology AMD licenses from nVidia and vice-versa (they both own a lot of patents on 3D acceleration... the IP was also the main reason why nVidia bought 3DFX, they never did anything with the actual technology... it was already outdated).
And let me point out, AMD's entire CPU division floats on the x86 license from Intel. They're completely at Intel's mercy there.
So nVidia has PhysX... Next time AMD might come up with something, which they can license back to nVidia.

 

[blastingcap]

"First to get caught." Interesting choice of words, but if this allegation is true, it is indeed serious. Even if NV was doing it first or second.

Thought the Cat AI thing was debunked as it doesn't appreciably deteriorate IQ? I have a Nikon D90 camera and I know it automatically tosses out some information when saving to file because that info is not detectable by the human eye. (Aside: NV recommends benching outliers like FC2 even for GTS450. How old is FC2 now? It's a blatant attempt to skew results, but I would not be surprised if AMD does this too with its Radeons.)

Vaporware is kinda typical of lots of companies, I won't say that's necessarily a sign of ill intent if ATI struggled to get physics support out the door in a timely manner. I'm not sure if they ever promised a concrete delivery date.

Anyway, you have made a decent case for why ATI hasn't been a sweetheart, either, but I think NV still has the darker history:

- bumpgate and NV's attempt to dodge blame for as long as it could.. $500 million to date just to deal with that disaster
- apparent illegal attempt at price-fixing to drive up costs of GPUs
- apparent sponsorship of marketing shills on AT forums
- NV also cheated on benchmarks before
- other stuff that I don't think rises to the level of the above but still annoy some people, like not enabling PhysX if ATI card is detected, or Batman:AA AA shenanigans, etc.

I think AMD made the wrong choice.

I won't quote your entire thing because this post is already long, but I can see where you are coming from. At the same time, your initial post made it sound like NV was doing ATI a favor by approaching them with PhysX, and that AMD was unreasonable to decline. I think it IS with the benefit of hindsight that you wrote what you did; perhaps at the time ATI had other ideas than to license technology from NV and to thus be at their mercy there. I think it's very arguable that turning down NV was the right thing for ATI to do at the time. That said, I don't necessarily disagree with your analysis. But at the time, based on its actions, ATI obviously thought it had better alternatives.

Speaking of bad business decisions, perhaps NV made the bigger mistake in recent years in not allying itself with Intel or AMD. It's now stuck in a position where it has no x86 license and its competitors are busy fusing CPU + GPU. The chipset business got hit first, but in addition to that, low-end discrete cards may be a distant memory soon. AMD could quite plausibly eventually figure out physics, but will NV eventually successfully deal with the x86 issue? I don't expect NV to go bankrupt or anything anytime soon, but it definitely has challenges to overcome. I want NV to survive and maintain competition with AMD so as to spur innovation and to keep a lid on GPU prices. Supercomputing and mobile are great and all, but I'm more interested in gaming. 😀

Well, for starters, ATi was the first vendor to get caught at cheating in benchmarks (the Quack.exe incident).
AMD is still 'half-cheating' in benchmarks because their AI feature still reduces texture quality as a trade-off for speed. AMD advises reviewers to test with AI enabled, but they do not specify what AI does exactly. Hence most reviewers aren't aware of the reduction of image quality and benchmark with AI enabled.

And should we get into AMD's OpenCL/physics practices? For years now, AMD has been making promises about supporting OpenCL, and delivering an open physics solution. So far, nothing has happened. They don't even include an OpenCL runtime in the end-user drivers.
They've rigged some Havok/OpenCL physics tests, showing 'working' physics acceleration, but somehow even years later, nothing has emerged (the earliest set up was a X1900+X1600 system... that was AGES ago. Where is it?).
So basically they've only deluded their customers, in an attempt to keep people away from nVidia (who DO have OpenCL and physics for everyone).
(If you thought nVidia showing a mock-up card at the Fermi introduction was bad... what do you think of this then?)

There's plenty more, but you get the idea. Bottom line is, AMD/ATi aren't exactly 'sweethearts' themselves.

Edit: you added stuff to your post which I will respond to here. AMD is at Intel's mercy, but AMD64 allows AMD room to live. Plus Intel is probably better off with a weak competitor.. it lets them point at AMD whenever anyone thinks about filing legal action against them for being a monopoly. Some consultants wrote a book recently about this... how it's better to be no. 1 and have a weak no. 2 survive and (barely) compete with you, so as to avoid bad publicity or anti-trust laws.

 

[Scali]

You added this edit after I responded to the rest of your post, so I want to say that I agree with the above to a large extent (definitely NOT if you are implying that AMD should make proprietary technology like NV does, though). AMD did introduce Eyefinity. It also pushed for tessellation first, but thanks to NV's influence, it wasn't included in DX10 but instead got pushed back to DX11. (AMD GPUs had tessellation since way back; in fact the console GPU (forgot what it was called, it's in xboxes) have tessellation hardware in them.

Incorrect.
Firstly, tessellation was already in DX9. It supported N-patches and RT-patches.
ATi supported N-patches with TruForm, nVidia supported RT-patches.
Since neither were a success, nVidia didn't pursue tessellation anymore at that time.

Secondly, ATi's tessellation in the XBox and DX10 cards was not much more than a glorified version of TruForm. The DX11 tessellation is completely different, and completely programmable, which makes it far more powerful (and might finally make tessellation a success).
It's not like AMD's DX10 cards can do anything even remotely similar to what DX11 cards can do with tessellation, so it's foolish to compare them.

 

[blastingcap]

Incorrect.
Firstly, tessellation was already in DX9. It supported N-patches and RT-patches.
ATi supported N-patches with TruForm, nVidia supported RT-patches.
Since neither were a success, nVidia didn't pursue tessellation anymore at that time.

Secondly, ATi's tessellation in the XBox and DX10 cards was not much more than a glorified version of TruForm. The DX11 tessellation is completely different, and completely programmable, which makes it far more powerful (and might finally make tessellation a success).
It's not like AMD's DX10 cards can do anything even remotely similar to what DX11 cards can do with tessellation, so it's foolish to compare them.

I am not saying they are technically at the same level! Just used it as an example of how ATI is not always a laggard. Just usually. 😉

 

[Scali]

"First to get caught." Interesting choice of words, but if this allegation is true, it is indeed serious. Even if NV was doing it first or second.

Ofcourse the 'allegation' is true.
Bothered to google for 'quack.exe'?
I think Hardocp was the first to bring this out in the open:
http://www.hardocp.com/article/2001/10/19/atis_radeon_8500_reviewed/5
3DCenter then did a deeper analysis:
http://alt.3dcenter.org/artikel/2001/10-24_a.php
Clearly reduced IQ.

And how is the choice of words 'interesting'? I'm just saying that ATi had its share of 'dubious' affairs, because some people think they're just the ideal company.
I'm not going to deny that others do it too. I know plenty of other examples... heck, even S3 and Intel have cheated in benchmarks. It's not just nVidia, but somehow AMD/ATi managed to brainwash the internet into thinking that.

Thought the Cat AI thing was debunked as it doesn't appreciably deteriorate IQ?

Doesn't matter, in my opinion. The driver does not do what the code tells it to do. It breaks API specs and its aim is to improve performance. I call that cheating, whether you can actually see the difference or not.
If an athlete uses doping, but isn't caught, did he not cheat then?

Anyway, you have made a decent case for why ATI hasn't been a sweetheart, either, but I think NV still has the darker history:

I really don't think that matters here.
If you want to be 'righteous' and avoid companies because of their dubious business practices, then you should avoid AMD/ATi as well.

As for bumpgate... Don't attribute to malice what you can attribute to incompetence (do you know how difficult it is to prove that such a failure exists and is indeed caused by a wrong choice of materials? It takes 2 years or more before the chips even start failing, and not all of them do. I lost an 8800GTS and a 9800GTX, but a 9800GTX+ still works).

I think it IS with the benefit of hindsight that you wrote what you did;

Then you'd be wrong. Because I've been saying the same thing from the beginning. If you bother to google forums, you will probably turn up my posts to prove this.

Speaking of bad business decisions, perhaps NV made the bigger mistake in recent years in not allying itself with Intel or AMD. It's now stuck in a position where it has no x86 license and its competitors are busy fusing CPU + GPU.

I think that's rubbish. nVidia never was a CPU company, so it never needed x86. Fusion is low-end rubbish, which nVidia will probably not be losing sleep over.
I think nVidia's Tegra move is good enough. The market for mobile devices is much larger than that for Fusion-based machines.
I think for the future, the threat of ARM-based CPUs moving into the x86-based laptop market is greater than that of x86-based CPUs moving into the ARM/Tegra-markets.

Edit: you added stuff to your post which I will respond to here. AMD is at Intel's mercy, but AMD64 allows AMD room to live.

Not really. AMD64 is just an extension of x86, and as such, Intel has more control over it than AMD does. If Intel decides that AMD did not honour the x86 license, Intel gets AMD64 'for free' (and speaking of 'goodness of heart', I think we can all thank Intel for not interpreting AMD's split of GlobalFoundries as a breach of that license, because they very well could).

 

[AnandThenMan]

Fact:
-Nvidia owns PhysX.
-AMD can buy license from Nvidia to run PhysX on their video cards.
-AMD have not brought the license from Nvidia to run PhysX on their video cards.

Can you point me to where Nvidia states they allow PhysX licensing on non Nvidia graphics cards? And how much they charge for this?

 

[ZimZum]

If Intel decides that AMD did not honour the x86 license, Intel gets AMD64 'for free' (and speaking of 'goodness of heart', I think we can all thank Intel for not interpreting AMD's split of GlobalFoundries as a breach of that license, because they very well could).

I doubt the Justice department would sit idly by and allow Intel to yank AMD's x86 license. Even if the wording of licensing agreement allowed them to do so. They already have an itchy trigger finger where Intel is concerned. These companies cant operate with impunity.

 

[blastingcap]

Interesting because of the interesting wording, which hedges against the possibility that NV was first but didn't get caught first, not that it matters--I am agreeing with you that it's shady. I also hedged by saying "if true" just in case there was someone prior to NV or ATI that cheated, known or unknown.

Yes, they are all businesses, and no, I don't think ATI/AMD are perfect, even if others think otherwise. Though I notice your lack of comment on NV's apparent attempt to price-fix + its viral marketing efforts on this board. Even without bumpgate, those two push NV in front of AMD in the "dubious corporate behavior" race so to speak. And INTC's payola scheme kind of speaks for itself as far as dubiousness.

I have never advocated the boycott of INTC, NVDA, AMD, or ATI and own hardware by all four companies. If I didn't need single-GPU 3-monitor I would have long since gotten a GTX460 1GB, despite my opinion of NV management. But I draw the line at things like bumpgate, because I want hardware to be reliable; I already explained why earlier and why I understand if others might think me paranoid or something for passing over a used 8800GT in favor of a stopgap AMD card to use for a year or two while bumpgate got sorted out.

"Incompetent" is the last word I'd use to describe NVDA.

No, I didn't google it up, I'm not that OCD about forum posts. It was still harsh to characterize AMD's rejection of the PhysX offer the way you did, given the position AMD would have been put in.

Tegra and supercomputing may ultimately end up a good move, even the latter, as it has far less competition than the former and potential first-mover advantage, even if the size of the market is much smaller. Sure mobile is a big market, but it's a crowded one. I've never disagreed with this and in fact got shouted down when I suggested the same thing in another thread a while back. However, I am more interested in PC gaming than in either mobile or supercomputing.

We shall see if the lack of x86 hurts NVDA in the end.

INTC is smart by keeping AMD barely-alive and allowing them GloFo. Keeping AMD around is (imperfect) insurance against DOJ anti-trust actions. "What do you mean we're a monopoly? We have competition!"

Ofcourse the 'allegation' is true.
Bothered to google for 'quack.exe'?
I think Hardocp was the first to bring this out in the open:
http://www.hardocp.com/article/2001/10/19/atis_radeon_8500_reviewed/5
3DCenter then did a deeper analysis:
http://alt.3dcenter.org/artikel/2001/10-24_a.php
Clearly reduced IQ.

And how is the choice of words 'interesting'? I'm just saying that ATi had its share of 'dubious' affairs, because some people think they're just the ideal company.
I'm not going to deny that others do it too. I know plenty of other examples... heck, even S3 and Intel have cheated in benchmarks. It's not just nVidia, but somehow AMD/ATi managed to brainwash the internet into thinking that.

Doesn't matter, in my opinion. The driver does not do what the code tells it to do. It breaks API specs and its aim is to improve performance. I call that cheating, whether you can actually see the difference or not.
If an athlete uses doping, but isn't caught, did he not cheat then?

I really don't think that matters here.
If you want to be 'righteous' and avoid companies because of their dubious business practices, then you should avoid AMD/ATi as well.

As for bumpgate... Don't attribute to malice what you can attribute to incompetence (do you know how difficult it is to prove that such a failure exists and is indeed caused by a wrong choice of materials? It takes 2 years or more before the chips even start failing, and not all of them do. I lost an 8800GTS and a 9800GTX, but a 9800GTX+ still works).

Then you'd be wrong. Because I've been saying the same thing from the beginning. If you bother to google forums, you will probably turn up my posts to prove this.

I think that's rubbish. nVidia never was a CPU company, so it never needed x86. Fusion is low-end rubbish, which nVidia will probably not be losing sleep over.
I think nVidia's Tegra move is good enough. The market for mobile devices is much larger than that for Fusion-based machines.
I think for the future, the threat of ARM-based CPUs moving into the x86-based laptop market is greater than that of x86-based CPUs moving into the ARM/Tegra-markets.

Not really. AMD64 is just an extension of x86, and as such, Intel has more control over it than AMD does. If Intel decides that AMD did not honour the x86 license, Intel gets AMD64 'for free' (and speaking of 'goodness of heart', I think we can all thank Intel for not interpreting AMD's split of GlobalFoundries as a breach of that license, because they very well could).

 

[AnandThenMan]

If Intel decides that AMD did not honour the x86 license, Intel gets AMD64 'for free' (and speaking of 'goodness of heart', I think we can all thank Intel for not interpreting AMD's split of GlobalFoundries as a breach of that license, because they very well could).

As part of the cash settlement with AMD, Intel has agreed that AMD can use GlobalFoundries, and is also free to use other foundries such as TSMC. (which AMD is using for their first Fusion products). Your "we can thank Intel" line is absurd and quite honestly ignorant.

And as mentioned, Intel is on a very short leash with the justice department.

 

[Pantalaimon]

You know, it's interesting, how those who defend NVDIA, seem to try to point out that the other side does it too, instead of trying to refute accusations leveled at NVIDIA, like the price fixing, bumbgate, and the puppy that is Fermi.

'The other side does it too' is not really a good defense when it comes to morally/ethically/legally questionable and wrong actions.

 

[Scali]

You know, it's interesting, how those who defend NVDIA, seem to try to point out that the other side does it too, instead of trying to refute accusations leveled at NVIDIA, like the price fixing, bumbgate, and the puppy that is Fermi.

'The other side does it too' is not really a good defense when it comes to morally/ethically/legally questionable and wrong actions.

If you are referring to me, then you got the wrong idea.
I'm not trying to excuse nVidia's behaviour... just pointing out that there are no 'holy' companies, so when people are going "I'm not buying nVidia because they did this-and-that", then I'm pointing out that by that same logic, they should not buy AMD either.

I have never excused nVidia or any other company for questionable actions.

 

[Scali]

Can you point me to where Nvidia states they allow PhysX licensing on non Nvidia graphics cards? And how much they charge for this?

I posted a link to that a while ago in this thread.

 

[Scali]

Interesting because of the interesting wording, which hedges against the possibility that NV was first but didn't get caught first, not that it matters--I am agreeing with you that it's shady. I also hedged by saying "if true" just in case there was someone prior to NV or ATI that cheated, known or unknown.

I don't think the wording is 'interesting', just that it is the only correct wording that I know of.
After all, ATi was caught in this instance. This may not have been the first time that ATi was cheating. ATi may not have been the only one cheating... ATi may not have been the first one that was cheating. We don't know any of these things for certain.
We only know that they got caught first.
And stop making this NV or ATi. Back then, there were more than just two players.
I know one of the earliest examples of cheating in benchmarks was S3 'magically' enabling their texture compression in applications. I also know that other vendors, after adopting their texture compression technology, also started to apply this cheat.

Yes, they are all businesses, and no, I don't think ATI/AMD are perfect, even if others think otherwise. Though I notice your lack of comment on NV's apparent attempt to price-fix + its viral marketing efforts on this board. Even without bumpgate, those two push NV in front of AMD in the "dubious corporate behavior" race so to speak. And INTC's payola scheme kind of speaks for itself as far as dubiousness.

What is there to comment on? What is understood does not need to be discussed.
And I'm not really interested in trying to 'determine' which company is 'most evil'.
I'm just saying that buying AMD because 'nVidia does nasty things' is silly, in the light of all the nasty things that AMD has done.
I think we should end it here, because you're getting too carried away. You're coming off as too pro-AMD for my liking, and you're being too aggressive, trying to attack me because you think I'm pro-nVidia. I am nothing of the sort.
You've reiterated that you don't like how I phrased things about 5 times now. I understood that the first time. I'm not going to change my words to please you. You should have figured it out by now. I have my integrity. I choose my words carefully, and I stick by them. I said what I wanted to say, how I wanted to say it. Now it is up to you to learn how to deal with that.

 

[Keysplayr]

You know, it's interesting, how those who defend NVDIA, seem to try to point out that the other side does it too, instead of trying to refute accusations leveled at NVIDIA, like the price fixing, bumbgate, and the puppy that is Fermi.

'The other side does it too' is not really a good defense when it comes to morally/ethically/legally questionable and wrong actions.

I believe it's necessary when those who try to bash Nvidia for it at the same time claim that AMD/ATI is an angel company. Or a saint. All things blastingcap listed above can be attributed to both companies, with the exception of bump issue. The problem is that he, and those like him, doesn't think AMD/ATI/Intel/OCZ/Samsung/Hitachi ad infinitum does the same. Or at least AMD/ATI.
As more time goes by, it seems that the more acts that were commited are denied they ever happened or tried to explain away somehow.
Whatever the case, both companies practice this, and the reason this needs to be mentioned is because it looks like a lot of AMD fans think otherwise.
Emphasis on the word "think".
Now you know why. You choose to think it is a defense for Nvidia. It isn't.
It's just keeping it real and a casting a little light on the subject.

 

[Keysplayr]

Yes, they are all businesses, and no, I don't think ATI/AMD are perfect, even if others think otherwise. Though I notice your lack of comment on NV's apparent attempt to price-fix + its viral marketing efforts on this board. Even without bumpgate, those two push NV in front of AMD in the "dubious corporate behavior" race so to speak. And INTC's payola scheme kind of speaks for itself as far as dubiousness.

Umm, it takes at least two entities to price fix. Who do you think the other was?
AMD uses viral marketing and was made public a few years ago. Known as High Road Communications. And yes, to think they aren't on this board is the stuff of pure fairytales. So, who's out in front now in the "dubious corporate behavior" race? Give me a break dude. You can't really be serious with this stuff at this point. No offense, but it's starting to become a little funny that you might actually really believe what your typing at this point.

 

[Lonyo]

I believe it's necessary when those who try to bash Nvidia for it at the same time claim that AMD/ATI is an angel company. Or a saint. All things blastingcap listed above can be attributed to both companies, with the exception of bump issue. The problem is that he, and those like him, doesn't think AMD/ATI/Intel/OCZ/Samsung/Hitachi ad infinitum does the same. Or at least AMD/ATI.
As more time goes by, it seems that the more acts that were commited are denied they ever happened or tried to explain away somehow.
Whatever the case, both companies practice this, and the reason this needs to be mentioned is because it looks like a lot of AMD fans think otherwise.
Emphasis on the word "think".
Now you know why. You choose to think it is a defense for Nvidia. It isn't.
It's just keeping it real and a casting a little light on the subject.

Most of it isn't even really relevant.
Cheating on benchmarks isn't doing anything for gaming, that's done for the company to put their products in a better light.
Quack/Quake didn't make people think "Oh man, I'm more/less interested in Quake" it made people think "This company is dirty and cheats in benchmarks".
Bumpgate has nothing to do with gaming. Same with price fixing.

Doing things "for gaming" would (historically) be things like adding new features (e.g extensions to DirectX like S3TC from S3 back in the day, or 3D features, or Eyefinity features), working to improve APIs, working to get features added.

In that sense, NV does a lot for themselves for gaming, but not for gaming as a whole mainly due to vendor lock-in.
People complain when certain consoles get things over other consoles or sooner (e.g. Xbox 360 exclusive bonuses vs PS3 exclusive bonuses). That's the sort of thing NV's approach results in, while they are free to "port" ATI work to their own system because it's openly available to do that (e.g. triple monitor gaming support, where ATI have worked with devs to get it implemented).

Historically I've switch GPU manufacturer almost every time I upgrade, so NV is doing nothing for me for gaming, because I'd only be able to use their features half the time, so it's not worth shopping specifically for those features. For people who stick only to NV they are doing a lot for their PC gaming.
For other people (people who stick to ATI), they are doing nothing for their PC gaming, and actually making it worse by locking them out of things.

 

[Keysplayr]

For people who stick only to NV they are doing a lot for their PC gaming.
For other people (people who stick to ATI), they are doing nothing for their PC gaming, and actually making it worse by locking them out of things.

Incorrect. Or at least I strongly disagree.
For people who stick to ATI, they don't "lose" anything, but remain status quo. Nothing at all becomes "worse".
For people who stick to NV, they benefit the extra. status quo + more.
In comparison, those who are interested in PhysX and 3DVision Surround for example, but own AMD hardware can perceive their situation as "worse" because they'd like to try it out. But they're really experiencing a given game as it was intended originally. With nothing extra. Sometimes "worse" without AA available, like in the case of Batman AA or SC2 initially. AMD NOW offers AA in both. Well, Batman AA collectors edition, and SC2. Nvidia is forcing them to do better, especially after the SC2 outcry.
Nvidia does far more for PC gaming. Without question. You choose to see it as a negative, which really does blow my mind sometimes with some of the "out of this world" comments that occasionally flash past my screen here. IMHO.

 

[Aristotelian]

You are the thrid poster I post this too directly, perhaps third time is the charm?

FarCry...AMD64 patch...look it up.

A) I wasn't talking to you in my original post.
B) Scali offered a list of transgressions therefore
C) I took him up on his kind offer.

He offered, I accepted, don't go telling people not to accept an offer especially when it is so kindly responded to as follows (rendering your post useless):

Well, for starters, ATi was the first vendor to get caught at cheating in benchmarks (the Quack.exe incident).
AMD is still 'half-cheating' in benchmarks because their AI feature still reduces texture quality as a trade-off for speed. AMD advises reviewers to test with AI enabled, but they do not specify what AI does exactly. Hence most reviewers aren't aware of the reduction of image quality and benchmark with AI enabled.

And should we get into AMD's OpenCL/physics practices? For years now, AMD has been making promises about supporting OpenCL, and delivering an open physics solution. So far, nothing has happened. They don't even include an OpenCL runtime in the end-user drivers.
They've rigged some Havok/OpenCL physics tests, showing 'working' physics acceleration, but somehow even years later, nothing has emerged (the earliest set up was a X1900+X1600 system... that was AGES ago. Where is it?).
So basically they've only deluded their customers, in an attempt to keep people away from nVidia (who DO have OpenCL and physics for everyone).
(If you thought nVidia showing a mock-up card at the Fermi introduction was bad... what do you think of this then?)

There's plenty more, but you get the idea. Bottom line is, AMD/ATi aren't exactly 'sweethearts' themselves.

Thanks for this. Perhaps we should create a separate topic about so-called ethical and unethical business practices by AMD and Nvidia. It'll be good for the general readership here to see that both companies have lipstick on their collars.

 

[Creig]

AMD uses viral marketing and was made public a few years ago. Known as High Road Communications. And yes, to think they aren't on this board is the stuff of pure fairytales.

Really? Could you please share with us the proof you are basing your accusation on? Because I would think that a Focus Group member would want to be very careful when accusing his sponsor's competitor of wrongdoing without some sort of concrete evidence.

 

[Keysplayr]

Really? Could you please share with us the proof you are basing your accusation on? Because I would think that a Focus Group member would want to be very careful when accusing his sponsor's competitor of wrongdoing without some sort of concrete evidence.

Who am I accusing specifically? And why should I be extra careful? I'm out in the open. And common sense dictates that the possibility of there being zero AMD affiliates or representatives on this board, is remote. You can argue that there is zero proof, but what does that really mean? It just means we don't know who they are. That's all.

 

[Grooveriding]

Who am I accusing specifically? And why should I be extra careful? I'm out in the open. And common sense dictates that the possibility of there being zero AMD affiliates or representatives on this board, is remote. You can argue that there is zero proof, but what does that really mean? It just means we don't know who they are. That's all.

So common sense dictates that we know there are still nvidia viral marketers on this forum ?

Because we know for a fact they had at least one here before, we even know who he was.

 

[Creig]

Who am I accusing specifically? And why should I be extra careful? I'm out in the open. And common sense dictates that the possibility of there being zero AMD affiliates or representatives on this board, is remote. You can argue that there is zero proof, but what does that really mean? It just means we don't know who they are. That's all.

You just accused AMD of having High Road viral marketers here at Anandtech, without providing any sort of proof to back up your allegation.

AMD uses viral marketing and was made public a few years ago. Known as High Road Communications. And yes, to think they aren't on this board is the stuff of pure fairytales.

Or did I somehow misunderstand your statement? If so, then I apologize.