3D Nature and Partners Declare ATI drivers Unsuitable for Professional Visualization

[HeXploiT]

May 4th, 2006, Morrison Colorado

3D Nature, a leader in photorealistic landscape visualization, together with a prominent group of 3D visualization developers and users, have taken the drastic step of publicly declaring ATI's Catalyst OpenGL display drivers as unsuitable for professional realtime visualization needs.

All past and current ATI Catalyst drivers for all ATI Radeon series display cards contain several critical bugs that render the hardware unusable in professional 3D visual simulation applications. ATI has been aware of the problems since 2005 but has not fixed the most critical issues, has not announced an ETA for fixing them, or even acknowledged them publicly in their Known Issues for current releases.

The two critical failures are referred to as "VSYNC spinlock" and "compressed SGIS_generate_mipmap failure". Both are documented extensively on the OpenSceneGraph web site:
http://www.openscenegraph.org/osgwiki/pmwiki.php/Tasks/OpenGLConformance
complete with screengrabs and sample code to reproduce the problem.

VSYNC spinlock refers to the situation where the display driver continuously runs the CPU during a wglSwapBuffers() function call when vertical retrace synchronization (VSYNC) is on. This leaves the CPU unavailable to perform other work during this period when it should normally be idle, drastically impairing the performance of modern multi-threaded visualization applications. Other modern display card hardware drivers (3D Labs, NVidia) leave the CPU available during this operation.

Continued...


Finally some pressure being exerted on Ati to fix the problem.:disgust:

 

[Keysplayr]

Well, hopefully they will now address it and get it fixed. Get ready for some spinnin.

 

[Gstanfor]

Finally someone from inside the professional 3D rendering industry has had the balls to come out and publically say what has been privately said for a long time now. Can't wait for joker, appopin and the rest of the fanatics to fly into full on denial over this, will have to check out Baumann's responses @ B3D too.

Come on ATi, just how long does an OpenGL rewrite actually take? You've officially been at since before R300 launched (just under 4 years now). Are your software engineers really that incompetent?

EDIY: It strikes me that the situation is made all the more amusing by ATi/the fanatics love of WHQL drivers and their berating of non WHQL drivers offered by the competition. Apparently only some certifications are deemed important...

 

[43st]

Thank goodness the Radeon series is for home use and not Professional Graphics. Which is basically the difference between rendering for speed (gaming) and rendering for accuracy (professional). If you're an IT manager and purchased a "home" product for a 3d department then you've made a poor decision.

I have a Radeon at home, I've also used Geforce, and use both FireGL and Quadro FX at work. Bringing a gaming card into the work place would be like using a Yugo to pull a tractor trailer.

 

[Gstanfor]

B3D link

Hi there, Chris from 3D Nature here. The guy that wrote the press release.

I guess you can imply that somehow our bunch of folks are clueless, or unprofessional. After all, I've only been working in 3D since about 1988, and OpenGL since 1996. But you probably shouldn't impinge on the reputation of people like Don Burns, Robert Osfield, Ben Discoe and Gordon Tomlinson. Don's been involved in GL since before it was Open, back at SGI, and is responsible from some important stuff that's still used in Performer today. Robert is the head of the OpenSceneGraph project, the most successful Open Source scenegraph on the planet, available for Windows, Linux and MacOSX. Gordon Tomlinson worked on Multigen-Paradigm's Vega scenegraph, and pretty much anyone who has ever done realtime CG terrain knows who Ben Discoe is. Together they've forgotten more about 3D than most people ever knew.

Most of us don't own ATI hardware anymore. We got smart. I'm stuck with one because when I bought my laptop, ATI was smacking NVidia around in the laptop market and everything you could buy was Radeon Mobility based. But it's not us we're concerned about. It's our users who get burned by this. They get sold a system with an ATI graphics card in it (probably a great value) and then they find out it doesn't work as it ought. Frequently, they're upset with us, the developer, because they think it's our fault -- because ATI couldn't possibly be neglecting serious bugs for so long, could they? So, we're sick of it. We tried asking, and then pressuring ATI into getting these issues resolved, and 6 months later, they're still broken. So, we decided we needed to publicly state our position and let ATI start taking the flack for it instead of us. We've told them for months we were going to go public with it if they didn't get it fixed, and here we are.

Yeah, Radeon isn't the top-of-the-line pro hardware from ATI. But that's not the point. NVidia cards that are comparable in price to the Radeon work just great. And, outside of a few rare exceptions, you don't find FireGL in laptops, period. Not everybody can justify a pro-level card anyway -- the goal of many of our companies is to bring professional 3D to the masses -- anyone heard of Google Earth?

As far a not using GL_SGIS_generate_mipmap -- GLU provides a (slower) software implementation for mipmap generation and always has -- and in fact most software automatically falls back to software mipmapping if the extension isn't available. The problem is the ATI drivers claim it IS available, so we try to use it, and find it's broken. If they'd just admit that it was broken and make the driver not offer the extension, everything would be great. If you're gonna claim to be able to do the job, then you better live up to the task, or go home. And, as noted elsewhere -- having a FireGL _doesn't_ solve the problem -- according to our tests GL_SGIS_generate_mipmap is broken there too!

Shrill? Yeah, I guess we come off a little shrill. We're sick of being ignored and the only way to get matters like this resolved is to make it apparent to ATI that there _is_ a downside to ignoring the problem. Otherwise they'll just keep ignoring it.

We do use the pro-level cards when the application calls for it. The latest quad Opteron to make an appearance in our company's office sports a pair of NVidia 7800GTX's.

I'm not an NVidia shill. I don't have any particular reason to pimp them, but at the moment, I _have_ to recommend their hardware to my users because ATI's isn't cutting it and 3D Labs is DOA. I'd _like_ to see ATI get this fixed because we find that our users (and therefore our company) _benefit_ from the competition between the manufacturers. ATI has a good price/performance ratio -- if the cards would just work right we'd be able to recommend them again.

 

[IlllI]

Originally posted by: Thera
Thank goodness the Radeon series is for home use and not Professional Graphics. Which is basically the difference between rendering for speed (gaming) and rendering for accuracy (professional). If you're an IT manager and purchased a "home" product for a 3d department then you've made a poor decision.

I have a Radeon at home, I've also used Geforce, and use both FireGL and Quadro FX at work. Bringing a gaming card into the work place would be like using a Yugo to pull a tractor trailer.

i was thinking pretty much the same thing

 

[Gstanfor]

clandren and Thera, by "professional" the author means "non gaming". Try reading the b3d thread or at least the author quote above.

 

[MercenaryForHire]

Hi there, Chris from 3D Nature here. The guy that wrote the press release ...

... And, outside of a few rare exceptions, you don't find FireGL in laptops, period. Not everybody can justify a pro-level card anyway -- the goal of many of our companies is to bring professional 3D to the masses -- anyone heard of Google Earth?

- M4H

 

[43st]

Originally posted by: Gstanfor
clandren and Thera, by "professional" the author means "non gaming". Try reading the b3d thread or at least the author quote above.

Google Earth works great with mid range ATI hardware last time I checked. Not to mention a 7800 isn't a professional card either.

Chris is having a tantrum and issued his "press release" to poke ATI with a stick. Last time I checked this didn't work too well. I've documented and reported my share of driver and software issues for some time to Autodesk. Sure it's a pain in the ass and sometimes a work-around is needed until they can address the issue. It's not a perfect world and developers, as any of us, don't have infinite resources to fix problems every time someone starts crying about it.

I guess I take issue more with his method than his accuracy. If Chris thinks he'll get support for is software faster this way then let him do it. But next time you need something from a man like this you'll probably think twice before using his service.

 

[xtknight]

So this does happen only in the Radeon series or does it also apply to FireGL? What I think they need to do is improve their lousy Linux drivers. NVIDIA's have always worked great for me but setting up my friend's 9800 XT with Xgl and composite extension was nothing but a thorn in the rear (the whole system froze every boot-up and we had to go into 'init 3'). Not to mention, the installer was also a pain, and poorly designed. But I guess that's not directly related to OpenGL. It is however something they 'said' they were going to do and hasn't been done. NVIDIA's installer is far from perfect either but it has had many less problems from my experience.

 

[43st]

BTW... if this is the quality of Open Scene Graph titles maybe they have bigger issues to deal with. I doubt pointing a finger at ATI is going to help them much... I guess we'll see.

 

[Praxis1452]

Originally posted by: Gstanfor
Finally someone from inside the professional 3D rendering industry has had the balls to come out and publically say what has been privately said for a long time now. Can't wait for joker, appopin and the rest of the fanatics to fly into full on denial over this, will have to check out Baumann's responses @ B3D too.

Come on ATi, just how long does an OpenGL rewrite actually take? You've officially been at since before R300 launched (just under 4 years now). Are your software engineers really that incompetent?

EDIY: It strikes me that the situation is made all the more amusing by ATi/the fanatics love of WHQL drivers and their berating of non WHQL drivers offered by the competition. Apparently only some certifications are deemed important...

Gstanfor why must you flamebait? Really before any ati "FanATIc" has even posted in this thread you must insult them first. Gstanfor your worse than any fanatic as shown in this post stop goddamn trolling.

 

[TanisHalfElven]

ATI's drivers suck. there hardware is probably the best on the market but there drivers pure suck. and the fact that so few games implement features ATI has like HDR+AA (oblivion)... and linux drivers suck even more. ATI will die if they don;t improve the drivers.

 

[Gstanfor]

some more clarification from the author

b3d link

Let me address a couple of items.

As far as work-arounds -- my application already has work-arounds. But, usually the work-around impairs performance. For example, my work around is simply to stop using compressed textures on Radeon series cards. Makes the end result look right, but greatly increases (I believe by a factor of 4) the amount of texture memory required by my application. So, (for texture usage anyway) a 256Mb Radeon performs like a 64Mb NVidia card. If _that_ doesn't inspire ATI to realize they're hurting themselves, I don't know what will. As far as the VSYNC spinlock, we have a predictive VSYNC loop that tries to estimate how long it will be until the VSYNC event, and avoids doing the wglSwapBuffers() call until we estimate there's only about 15% of the frame timeslot remaining. That way, instead of wasting _all_ of the unallocated frametime in busy looping, we keep our other work threads doing useful tasks longer, and only waste 15% of the time. Why 15%? because with CPU loads and program uncertainty you don't want to overshoot your estimate (causing a lurch in the redraw), so instead you are forced to undershoot to be safe. So, our application can do 15% less work on an ATI system, resulting in 15% worse performance than on a comparable system with a properly working VSYNC -- like an NVidia or 3DLabs card.

As far as professional goes, I think you set the bar too high. A system doesn't have to be a 6-headed flight simulator installation to run professional software. A huge number of our users (mine and our listed partners) use (and are only allowed to use) completely stock-standard business PCs. If they're US Government (as many are), they often are _required_ by purchasing to buy from a pre-established vendor, often someone like Dell. So, they're out there in the Real World, trying to do Real Work, using something that's probably less powerful than your average enthusiast gaming machine.

And that's how it works. We deploy top-of-the-line systems whenever it's called for, and the customer has the choice. But we also have to make do with what the customer brings to the table.

digitalwanderer>but what you're basically saying is that your non-ATi drivers for an enthusiast level card aren't capable of doing what their professional line-up can do and you're miffed about it....right?

Actually, what I'm saying is that my ATI card, regardless of what it was sold as, doesn't live up to its specs. If you claim to implement an extension, then it better work. If not, don't offer it as implemented. It so happens that software with large datasets (like ours) are more likely to utilize SGIS-mipmap and compression, but no where in the GL specs does it say "pro-only, don't let consumers have power like this!". It's just a common extension, and everyone and their brother has offered it for forever.

http://oss.sgi.com/projects/ogl-samp...ate_mipmap.txt

Likewise, nowhere does the wglSwapBuffers() documentation say "only professional cards should do this efficiently, make sure consumer-level cards just waste the CPU time".
http://msdn.microsoft.com/library/de...pnglr_1ss3.asp
But it's sort of implied that a modern graphics system doesn't spin the CPU idly when doing _anything_. Heck, we had sleep-till-interrupt double-buffer-swap-sync capability on the Amiga platform in, what, 1986?

There's nothing pro-only about these features. The only reason we qualified our announcement with "pro" is because we wanted to avoid the obvious (and inapplicable) arguments like "well, Quake runs fine on it, so it must be your fault". Yes, I do feel that Google Earth is the exact epitome of the pro-level application (formerly known as Keyhole) becoming the ultimate consumer application. I know it has workarounds of the same type as my own software. I know those poor guys over at Google Earth must tear their hair out daily trying to ensure that that program works on all these crazy computers and graphics systems around this planet. GE works in both DirectX and OpenGL, which simply doubles the number of bugs they have to cope with. I don't envy them, but they have a few billion dollars so they can throw more engineers at the problem. Wouldn't it be great if we could just get the vendors to take responsibility for their bugs? It's a dream world, I know, but thow much more time would ISVs have to make great software even better if we _each_ didn't have to fight with working around the same bugs over and over?

Finally, yes, as I said, there are various work-arounds available, with differing degrees of success. The easiest work-around is not to buy the ATI hardware in the first place. But, in the end, which do you think is a better end-result:
One company, ATI, never fixes the bugs and hundreds if not thousands of OpenGL programmers are forced to find and implement semi-effective workarounds to deal with them. Or,
One company, ATI, fixes their bug, and thousands of programs on millions of computers magically start working better.

As far as we can tell, yes these features have always been broken. Why is it an issue now? Because as we continue to push the envelope, we need features like this to work. Back in 1999 or even 2003, maybe spinlooping during vsync could be overlooked. But today, with multiple CPUs and cores, and lots of threads running trying to get the most out of every system, it's inexcusable. During my time working with OpenGL I've had to pressure every single OGL IHV into fixing bugs, S3, NVidia, 3DLabs, Matrox, Dynamic Pictures (later bought by 3DL), don't even get me started on that Intel Extreme thing. IHVs only fix bugs when there's motivation to -- bug fixes don't sell new cards, whiffy new features do. Developers _have_ to do this to keep them honest. As someone else noted, they fix game bugs a lot faster, because they're a mass-market lots-of-ticked-off-users problem. Those of us will higher-end niche products must find our own ways of raising awareness.

ATI doesn't owe me anything, and far from what was implied by Demirug, we shouldn't have to somehow bribe ATI into fixing their bugs. We have gone out of our way to make ATI aware of the problem, provided screenshots, source, executables and data to replicate the problem. It's not a problem of us misusing their driver and we need their help to set ourselves straight. It's their problem and we've done everything in our power to help them be able to fix it. I've been the point of contact for 6 months of back & forth with them about getting this fixed. Our business is to write our software, not make and give away demos in order to somehow persuade IHVs into fixing bugs. As we've said before, and the whole point of this release, if they're not gonna fix them, we're simply going to stop using their hardware, and tell people why we're doing so.

We mentioned the Radeon series in the PR because that's what people are familiar with -- we didn't want this written off as "oh, it's only some exotic card that's broken". I suppose for completeness we should have said that FireGL is busted too. But, I know the mipmap feature is broken, and I believe (though it's not listed on the OSG page) I think vsync is too. I don't have a FireGL to try it myself, but I've heard plenty of griping from people who did buy them.

Vysez>Why should Ati provide support for something that they never advertised in the first place?
I guess ATI never specifically listed SGIS_generate_mipmap as a feature in their magazine ads -- once you put in all the explosions and Dawn & Dusk's curvy... faces, there's not enough room. But one could argue that when they list SGIS_generate_mipmap in the glextensions call, they're advertising that it's there and can be expected to work. As I said before, if you can't make it work, don't have your driver offer up a busted implementation that software will be convinced to try and use.

Humus, thanks for pointing out that the sample model for the mipmap cause is a broken link. It doesn't actually matter what model you use -- converting any textured object using compression will cause the problem -- this was just the model used for the screenshot. I'll revise the web page to note this.

And as far as us doing this for our own benefit -- what benefit, exactly, do you think I'm getting here? I have spent uncounted days of my time tracking down the issue, replicating it, documenting it and communicating it. When all that failed, we wrote and reviewed a factual but firm announcement and got it out here. And, for our efforts, we enjoy getting scragged on on forums across the Internet. All, for the sole result of... helping get ATI's drivers working better for everyone? Man, I love my job. If you think my company, or any of our companies, is reaping great rewards somehow from this, I have some dot-coms you might want to invest in. I guess the only folks who are benefitting from this are probably NVidia. If I were an NVidia plant, I guess I'd have done my job well, but I'm not that devious of a person.

To close, I will pass along a quote of a quote of a quote of an e-mail that someone sent me last night:

Quote:
From: Essa Qaqish
To: Glenn Patterson; Toshiyuki Okumura; Larry McIntosh
Sent: Fri May 05 16:28:45 2006
Subject: RE: 3D Nature and Partners Declare ATI drivers Unsuitable for Professional Visualization

All,
Unfortunately, this is true and we are currently aware of it. There is an EPR and a committed fix scheduled for 8.261 (June posting).

While we can't tell from this when the fix was made, and thus if it was motivated by our wheel squeaking or not, it does confirm that it is a real issue and just maybe, after 7 months, we might see a fix for it. And we can all be happy. And all of this can fade into history. And that's all we've asked for.

 

[akugami]

For all those taking the time to bash ATI's drivers wholesale, I can tell you right now that ATI's Catalyst drivers are relatively stable. Is it bug free? No, but neither are nVidia's drivers. I've read about nearly as much nVidia bugs in it's drivers as ATI's drivers. I have read about people on this very forum (not to mention other forums) who have had to go back to older nVidia drivers because the newer ones were buggy. And these are knowledgeable folks who do the whole Driver Cleaner thing. So please stop with this tired, and partially untrue, snipe at ATI for having bad drivers. That said, ATI does have problems with it's Linux drivers and by most accounts, it's a pain in the rear.

Next up, are these driver bugs only in the Radeon series of cards? Or are they in the FireGL drivers and cards as well? If they're only in the consumer Radeon cards then part of the problem is overblown as you should not expect pro level precision and features in a consumer card. That is not to say that ATI should not be more responsive and fix the bugs, or at least disable features it does not actually support in the driver level so that the bugs to do not crop up. If they were already implementing a fix, ATI should be better at communicating that to developers and is entirely the fault of ATI.

 

[Canterwood]

Originally posted by: Praxis1452
Gstanfor why must you flamebait? Really before any ati "FanATIc" has even posted in this thread you must insult them first. Gstanfor your worse than any fanatic as shown in this post stop goddamn trolling.

Agreed.

I own Nvidia harware, and I still get tired of this tool trolling.
Dude, get a life and maybe you'll get laid or something!!!

 

[Keysplayr]

Originally posted by: Thera
Thank goodness the Radeon series is for home use and not Professional Graphics. Which is basically the difference between rendering for speed (gaming) and rendering for accuracy (professional). If you're an IT manager and purchased a "home" product for a 3d department then you've made a poor decision.

I have a Radeon at home, I've also used Geforce, and use both FireGL and Quadro FX at work. Bringing a gaming card into the work place would be like using a Yugo to pull a tractor trailer.

I believe they were talking about professional graphics use. E.G. FireGL (rebranded Radeons) and Quadro (rebranded GeForce's).

 

[BenSkywalker]

Bringing a gaming card into the work place would be like using a Yugo to pull a tractor trailer.

And nVidia's Yugos somehow manage to do it just fine......

This problem effects both FireGL and Radeon- read his comments.

 

[ronnn]

Originally posted by: keysplayr2003


I believe they were talking about professional graphics use. E.G. FireGL (rebranded Radeons) and Quadro (rebranded GeForce's).

I read it has being radeon under discussion, as firegl only has only one rarely incountered bug. But what do I know, if it runs Quake4, I am happy with ogl performance.

 

[BFG10K]

Here's hoping ATi get things fixed for the sake of the professional market.

Fortunately their drivers are superb in consumer (gaming) space.

 

[Wreckage]

Originally posted by: akugami
For all those taking the time to bash ATI's drivers wholesale, I can tell you right now that ATI's Catalyst drivers are relatively stable.

Cleary the professional community thinks otherwise.

Is it bug free? No, but neither are nVidia's drivers. I've read about nearly as much nVidia bugs in it's drivers as ATI's drivers. I have read about people on this very forum (not to mention other forums) who have had to go back to older nVidia drivers because the newer ones were buggy. And these are knowledgeable folks who do the whole Driver Cleaner thing. So please stop with this tired, and partially untrue, snipe at ATI for having bad drivers. That said, ATI does have problems with it's Linux drivers and by most accounts, it's a pain in the rear.

Calling out a problem is not an attack on ATI. Very few people who have used both would say ATI has a more reliable driver than NVIDIA. I know you feel the need to defend ATI (and yet claim no bias at all), but this is a problem that should be addressed and discussed.

Next up, are these driver bugs only in the Radeon series of cards? Or are they in the FireGL drivers and cards as well? If they're only in the consumer Radeon cards then part of the problem is overblown as you should not expect pro level precision and features in a consumer card. That is not to say that ATI should not be more responsive and fix the bugs, or at least disable features it does not actually support in the driver level so that the bugs to do not crop up. If they were already implementing a fix, ATI should be better at communicating that to developers and is entirely the fault of ATI.

Many people use professional packages on standard hardware. You don't always need a high end FireGL or Quadro card to do this kind of work. It was also noted that standard NVIDIA cards run these programs just fine, so you can't write it off as they are using the wrong kind of card.

ATI has done a good job lately with its Catalyst drivers. Now they need to focus on the OpenGL part. A full rewrite could solve not only this problem, but their problems with OpenGL games and Linux.

 

[nitromullet]

It's interesting to me that these guys would use Omega Cat 5.10a drivers for this article. You would think that if they are trying to make an appeal to ATI to have them address issues in their drivers that at least they would support their argument by showing examples using a current stock driver.

 

[akugami]

Wreckage, false generalizations at every turn is an attack. The truth is ATI has drivers that are about on par with nVidia as far as stability goes in the Windows front. Yet at every turn some fanboy is there saying ATI drivers are horrible. When you make false generaliztions like that, it is an attack. If I did that about nVidia drivers you'd be screaming bloody murder. Heck, when you consider the fact that you read about as much nVidia driver problems as ATI driver problems, if ATI has horrible drivers then so does nVidia. Next time I see someone post generalizations about ATI's horrible drivers I'm going to do that.

As for using nVidia drivers, I have a dual monitor setup. Currently one monitor is devoted to my brother's 7900GT SLI setup which I'm borrowing to look at the differences between the 7900's and my X1900's. I also help troubleshoot computers for my friends who game and some have ATI and some have nVidia cards. So I am a much better judge of the differences between the two architectures than someone like yourself who I doubt has even used a system with an ATI card except in passing.

And anyone who expects professional level features and service from a consumer level card is an idiot. If it has pro features on a consumer product, great. More bang for the buck. If it doesn't, well, it's not a pro product. Just like I wouldn't expect RAW support and changeable lenses on a consumer digital camera unlike a prosumer or professional one.

 

[Gstanfor]

Originally posted by: nitromullet
It's interesting to me that these guys would use Omega Cat 5.10a drivers for this article. You would think that if they are trying to make an appeal to ATI to have them address issues in their drivers that at least they would support their argument by showing examples using a current stock driver.

Most people wouldn't have to use the Omega drivers. The author of the article is an exception though because he uses a notebook pc with an ATi graphics chip in it - a graphics chip ATi will not support with drivers! So he is *forced* into using Omega drivers. Just for reference a large proportion of the leaked non beta Forceware drivers that appear are written for notebooks with nVIDIA GPU's. The inf's are then modded to extend support to all nVIDIA GPU's...

 

[BFG10K]

Very few people who have used both would say ATI has a more reliable driver than NVIDIA.

Perhaps; but I would be one of them.