Slashdot Mirror
Anand Tours ATI and NVIDIA
logicalstack writes "The folks over at AnandTech have written an
expose on their visits
to both ATI and NVIDIA. Interestingly enough ATI's facility shrouded in secrecy and NVIDIA's is quite open, Including full color pictures of their server farm, and a pic of the NV30 test machine the 'Ikos.' The CEO even showed off the old school NV1 with 1MB of ram!"
Here
I've had enough abrasive sigs. Kittens are cute and fuzzy.
Direct link to the article
NVIDIA programs their GPUs into the IKOS boxes and they run what is effectively a very large NV30 at very slow speeds. Very cool and very expensive.
When I was there a few years ago, they would sometimes hijack all the desktops in the company for more power. If I remember correctly, they would boot them into linux at night and make (slashdot cliche imminent!) a beowulf cluster!
-Greg Daly, formerly of riva extreme, aka
Try reading the article:
ATI imposed very strict restrictions on photographs during our visit to their offices in Thornhill, Ontario; we saw a lot of interesting things at ATI's offices (including the foundation for their fountain of fire in the lobby of their main building) but we weren't able to take pictures of most of them. On the other hand, ATI sat us down with one of their chip architects and we were able to get a wealth of information about how their GPUs were made.
NVIDIA wasn't able to set us up with any engineers for an extended period of time (although lunch with Chief Scientist, David Kirk is always informative) but they were much more lax on the picture front so we were able to bring you more of the behind the scenes from NVIDIA.
ATI just didn't want anybody taking pictures, but they were the one sharing the real information.
Never underestimate the bandwidth of a 747 filled with CD-ROMs.
Hear hear! I refuse to use Nvidia based on that fact. I have an ATI Radeon 7200. It may not be as fast, nor may it be quite as good, but at least I'm not putting proprietary software on my machine when there is an alternative.
They're binary-only because they don't own all the code used in them so it would break other licenses to publish it.
I use those binary-only drivers myself with a GF3 and have had no problems with X crashing.
ATI was the company that provided the in depth talk with a chip engineer. With NVIDIA, Andtech had to settle to having lunch with their lead architect. NVIDIA was okay with pictures, but ATI was the one that provided real information.
"Old man yells at systemd"
Exactly. I was quite fucking floored after the claim that ATI was 'shrouded in secrecy." They let a bunch of fucking people tour their facility. How the FUCK is this shit fucking shrouded?! Sounds pretty fucking open to me. Sounds like the fucking opposite of shrouded.
The picture thing I can understand. Maybe the intern has some saucy shit up on the screen and didn't read the motherfucking memo close enough to note that some fucking strangers were walking around the place ready to snap a billion digi cam photos.
Fucking christ on a moped: who give a fuck what the nVidia server farm looks like? I don't recall buying a video card based upon the size and configuration of some fucking SERVERS.
Perhaps he uses it to play Billardgl? Perhaps the free part of Tux Racer.
I know, I bet it was to play Tux Kart
Hrm... Maybe we can't assume he's playing closed games?
Well, I suppose we could go over the reasons for the billionth time on /., but there's no reason to believe that you'll listen this time either.
Oh well. Here it goes anyway.
The primary reason is that they cannot. They do not own all of the code that is in the drivers. There are extensive cross-licensing agreements between nVidia and SGI, dating back to the creation of nVidia from a bunch of ex-SGI engineers and the ensuing lawsuits. A good bit of the core code in nVidia drivers is essentially owned by SGI. nVidia cannot release that code. Period. End of story.
The secondary reason is that there is reason to believe that there are trade secrets in the drivers. Why do most people still favor nVidia over ATI? Because of the drivers. They work damn well most of the time, and the drivers you download today still support the original TNT. Additionally look at the GeForce4 and the Radeon 8500. On paper the Radeon 8500 was superior, and yet the GF4 beat it in benchmarks consistently. Why? The drivers. They were more mature, better written, and streamlined.
Don't like the situation? Fine, don't buy an nVidia card. What? Nobody else has 3d acceleration worth a crap? All the other drivers are just as unstable and slower too? Well, gee, maybe there's more proof that nVidia knows what the hell it's doing. Yes, it sucks if you're a *BSD fan or something else such that the binary-only drivers aren't usable, but, again, nobody made you buy nVidia.
Frankly, nVidia has spectacular Linux support. They release the Linux drivers within weeks of the Windows drivers and they're pretty damn stable (frankly, I suspect that if you have continual issues here that it's some other piece of hardware being marginal and pushed over the limit by running the card at full functionality). Oh yeah, and they're fully functional... don't forget that little bit.
It's really sad to see people whining for Linux support, getting pretty damn spectacular support, and then whining that it's not good enough. No wonder most manufacturers don't bother - damned if you do, damned if you don't. So why spend the time and money on a marginal market if you're just going to get roasted anyway?
$20 says you'll use your ATI card to play Unreal Tournament 2003
On Microsoft Windows, of course. 'Cuz last I heard, the ATI cards won't work for UT2003 on Linux. Mind you, this may have changed, and I speak not from experience as I still have a shitty Rage 128 in my machine (hey, it handles Q3 and RtCW fairly well).
I'm trying to decide between upgrading to a Radeon 8500 or a GF4-Ti4200, and I'm leaning toward the GF4 because I'd rather have a proprietary binary driver that works than an open-source one that doesn't. So if anyone has had a good experience running a Radeon 8500 under Linux (especially with UT2003 demo, either the open-source dri driver or ATI's binary release), please let me know.
Hey kids, there's only 5 days left 'til Yak Shaving Day!
Define up-to-date?
Do you mean, same gameplay, more realistic graphics?
Please, (and I really mean this) Gaming industry, give us some kind of new game. It's just the same crap over and over and over. Like TV Sitcom's the games produced these days are stuck in a rut.
> (and buggy) Linux drivers?
You might want to try switching from nvagp to agpgart, or vice versa, depending on your mobo.
I went from agpgart->nvagp a few months ago, and suddenly most of my stability problems with the drivers were gone.
But they work find for the majority of the people that use them. I'm not saying their drivers are perfect, but I'd be willing to bet that the nearly all the people complaining of them crashing are having problems with something else.
I'm glad you could publish your drivers under the GPL, but not everyone is and I'd rather have closed drivers that work well for me than no drivers at all.
> Why can't nvidia at least do this for its older cards.
Because NVIDIA has a unified driver architecture. ATI is trying to do this too now. Actually, they started doing it with the 8500, but they rewrote the drivers for the R200, and broke 8500 compatability, which means unified drivers for anything past the R200.
actually, they said they had lunch with the CEO. Who I might add, works in a cubicle like the other employees. Wanna compare/contrast with ATI's CEO?
Liberty.
Once again, FUD... I've posted this before, and I'll post it again:
SGI has stated, on a number of occaisions, that they are not responsible for the closed nature of nVidia's driver and that they have, in fact, tried to push nVidia into opening the drivers...
Please learn all the facts before posting that crap again.
Dinivin
I fully agree. Last year I was at GDCE doing research at tying to get info from both ATI and nVidia for articles I was writing. all the people from ATI I met were fantastic, I spoke to a few of the heavy engineers (huge kudos to Alex Vlachos and Jason Mitchell) along with the PR and Product Management people. They all went massively went out of their way to help me and inform me, answering any questions I had, burning me cd's of demos, pics, info, etc and following up further technical questions via e-mails and phone calls after the conference was over. This was the overall attitude of ATI at the conference.
Next we come to nVidia, I repeatedly came up against a brick wall, the case was the same for other developers, with David Kirk doing a fine politician-style non-answering of questions after his presentation. You generally got the impression that there were a select few that may be lucky enough to be given certain information, but it was very much on nVidia's terms.
Fair enough, companies have secrets which they need to keep, but from my experiences with the companies, ATI are far far more open. If anything this article backs that up. Would you prefer a bundle of photos or a chance to talk with a variety of the actual engineers?
I believe the statment really goes:
Linux is for those who hate Microsoft, BSD is for those who love UNIX.
Considering that BSD origonates from the origional UNIX, that makes more sense...but then again you are probably a linux zealot who thinks RMS is some sort of man god...
I don't think SGI's old IP is the reason NVIDIA won't release source code. The real reason is that the drivers deliberately cripple certain advanced features on NVIDIA's low-end cards, to force "professional" users to buy their high-end cards.
Remember how the $600 Quadro2 hardware was exactly the same as a $200 GeForce2, except for a tiny little resistor? I'm sure there are a few places in the NVIDIA driver like:
if(user_paid_for_quadro()) {
make_antialiasing_fast();
enable_overlay_planes();
} else {
make_antialiasing_slow();
disable_overlay_planes();
}
So naturally a few days after they release the driver souce, somebody would provide a "magic" version of the driver that makes all of NVIDIA's low-end cards perform just like their high-end cards. Then they wouldn't be able to charge $600 for "pro" video cards anymore...
This is certainly true for older hardware (pre GF4), and is again an artifact of the SGI cross-licensing with nVidia.
The GF4 chip is separate from the GF4 Quadro chip though - at least as far as pinouts go (it may be that the actual core is the same still, but fat lot of good that does if there aren't leads for the "professional" bits). The GF2/Q2 chips and cores were identical excepting a resistor, as you note.
There are also (allegedly, I certainly have not confirmed this) SGI-only features in the core. I can't imagine that these functions would be exposed at all in public drivers though, so I can't see that being an issue.
the trade secret angle is, of course, completly off.
How can it be a trade secret when it will only work with there card? It's not like ATI is making the same chipset. If ATI wanted to know how there drivers work, they would know. Hell, I'd be surprised if the don't have a de-compiled version on there systems now.
That said, I am pleased with nVidia support. I even wrote them some letters when they first started doing it, to show my support.
The Kruger Dunning explains most post on
Hear hear! And why walk with *two* legs when we clearly only need one!
I don't get it. I'm all for Open Source, but I'm even *more* for a company taking an active interest in supporting their hardware under Linux. I've got a GEForce2 on my system and the drivers are *sweet*. Full support of *all* the hardwares features. How often do you get that under Linux? Not to mention the fact that the drivers compare nicely with their Windows counterparts.
Why spend the same amount of money for hardware that has less support and will effectively run slower because of it? I just don't get it...
If every hardware company were like NVidia we would have far less trouble buying a new printer/modem/videocard/etc.
"Ignorance more frequently begets confidence than does knowledge"
- Charles Darwin
Everybody (up to the CEO) at ATI works in cubicles, as well. It's actually quite common in most techie companies.
Never underestimate the bandwidth of a 747 filled with CD-ROMs.
Gentoo has a good page filled with information about NvAGP vs. AGPGart here.
The page isn't Gentoo specific (I use debian personally)
-- Bob
1^2=1; (-1)^2=1; 1^2=(-1)^2; 1=-1; 1=0.
NVIDIA is like a breath of fresh air when it comes to corp policy
That explains their stock price...$8.95 down from a 52-week high of $72.66. Of course, part of that is that Microsoft changed the Xbox chips, leaving NVidia holding the bag (or goodwill as they say in the industry). I don't quite see how yet another company getting screwed by Microsoft qualifies as a breath of fresh air, but what the hell.
Not that ATI stock is anything to write home about -- down about 50% on the past 5 years. I attribute that to some of the worst drivers knows to man. ATI up until about six months ago reminds me of Apple in the Amelio era.
For the record my computer has a nforce chipset and a radeon video card. And it runs windows just fine!
With ATI, they answered my questions accurately but briefly, and made no particular response to the bug reports I posted (which finally seem to be fixed now - mostly - in 02.3, 8 months after I notified them). Their dev support team gave me adequate support, IMHO. Their regular user technical support was not even close to adequate, the once or twice I've used them - I had to track down the fix myself and explain it to them (which they never even acknowledged).
nVidia OTOH went out of their way to explain their extensions, listen to my suggestions, meet with me personally etc etc. "Outstanding support" would be a better description - they told me all I wanted to know (within reason), and I felt like they were listening carefully to the suggestions I had for future hardware. Haven't tried their ordinary tech support, if they actually have any (being a chip maker not a board maker like ATI).
Why would anyone engrave "Elbereth"?
I mean up-to-date in terms of gameplay, graphics, and depth. While tuxracer is a decent distraction, we're not talking about something I would drop money on a new card for, or even something that would have been a marketable, high dollar ($50, instead of say, $10) game 4 years ago.
The unfortunate fact of the matter is that even if the gaming industry rehashes gaming concepts, they have a level of polish that can't be found in open source games. Take, for instance, Deus Ex. Was it revolutionary? No, it was a first person shooter. Was it highly interactive, engaging, and immersive? I think so. Were the production values high? Yes. And that is what is missing from the current line-up of open source games.
Open Source development cannot hope to keep pace with commercial game development. The time frames are too short, production costs too high, and talent at too much of a premium to create games that would require a high end graphics card. Which is what we are talking about, right?
Some types of games are very successful in Open Source (See: mahjongg, which I play every day). But the games that use a high level ATI or Nvidia card are not developed by the Open Source community. Engines, maybe (see Id), but the game content, no way. Maybe someday, but not today.
*everything* is Orwellian to cats.
It's not nearly that simple.
NVIDIA cards are unlike anything you've ever seen on the inside. It's not a simple matter of register banging like most hardware. And yes, there is quite a bit of proprietary/trade secret stuff in there, such that publishing the driver source or opening the hardware interface would be detrimental to NVIDIA.
As much as we all hate it, the tech industry is largely driven by trade secrets, patents, and lawsuits. I don't think anyone at NVIDIA really likes that, but it's the only way to survive given the broken state of IP laws in the world.
-John (having contributed to the driver)
But you'll believe the guy who said that the problem is SGI IP without providing any links? Says a lot about you.
Dinivin
None of what you've mentioned has been advancing gaming far. I'll give you, the increase in immersive feeling has done some for gaming, but we are reaching a limit. Increasing the look and feel only goes so far. Can you tell me that you've felt the gaming pull as hard as you used to?
I think that Free games will eventually max hardware out, but you have to remember that growth in Free Software works slower. The difference is that it is steady, and eventually passes propreitary software, which has usually moved on and is looking for profit somewhere else (look at mozilla).
My original point is that a person who is dedicated to using a 3D card in Linux in a non-propreitary manner can still get games to play that are decent.
The original poster implied this wasn't true.
I think the next revolution needs to be in free content creation. This is he only way that things will move forward.
I think people need to be more creative.
I think part of the reason that the DMCA dosen't get people's ire up anymore is that most people don't create content. That needs to change. We used to be a more creative world. This will help the free gaming cause.
I can see your points.
As an aside: You may want to checkout billardgl, a good use of graphics (and uses a decent amount of the graphic card).
So, any suggestions about what the best, cheap upgrade car for a two year old Mac would be? It's not worth it to me to shell out $200 or more for the top of the line hardware -- I don't play video games or anything like that -- but if a video card in the say $50 to $75 range would give a noticeable boost then it might be worthwhile.
Does anyone know what the minimum video hardware is to get QE running and how much it would cost to get that hardware costs these days? On the same lines, given similar hardware, have people seen better gains by upgrading graphics hardware or adding more ram? For the money I'm willing to spend right now, I could throw in half a gig of ram, but I've heard that upgrading the video card could be almost as much of a performance boost. It would be nice to get a few more opinions on which upgrade path makes more sense...
Thanks!
DO NOT LEAVE IT IS NOT REAL
Funny, you've been asked twice now and declined to provide links.
Here's mine:
FreeBSD Driver Initative
Announcement of collaboration between NVIDIA, SGI, and VA Linux
NVIDIA press release
And another release
Tom's Hardware discussion
Oh, and SGI isn't the only proprietary code either. There's also a cross licensing agreement with S3 for the S3TC (S3 Texture Compression) algorithms that NVIDIA doesn't have the right to disclose.
NVIDIA and SGI drop lawsuits
> 9. A little something called XBox.
Wouldn't "A big something" be more accurate?
Of all those links, only two talk about how SGI's IP is preventing nVidia from open sourcing the drivers. Neither of those two (Tom's Hardware and the FreeBSD nVidia driver initiative) can even remotely be considered official statements by either nVidia or SGI.
Care to try again?
Hey, I'll gladly admit that I have no proof that my statement is true. I have simply based it on what an SGI employee has told me.
How about you admitting you have no proof to back up your statement about SGI's IP.
Oh, and your final link should read "NVIDIA and S3 drop lawsuits" not "NVIDIA and SGI..."
Dinivin
LSB isn't a screwy kludge. Lots of other people did it as well. Its not exactly as if one is right and the other is wrong. Besides, PPC can be programmed into either big or little endian mode. Its Apple's fault for not being industry standard.
See the Endian FAQ
A deep unwavering belief is a sure sign you're missing something...
Because
A) There is licensed stuff in the GL implementation that they can't release.
B) ATI's major weak point is drivers. Since OpenGL drivers implement the *entire* OpenGL API, rather than just bang hardware like other drivers, opening the code would hand over a *lot* of optimizations to ATI. Helping your competitor out with their one weak aspect is asking a bit too much of a company.
And NVIDIA's drivers a rock-solid for me an many others. I've used them on a Riva TNT1, GeForce2 MX, and GeForce 4 Go 440. They've worked *perfectly* on every single one. I haven't ever had an X crash. Now, if you're having problems, that's just your setup.
A deep unwavering belief is a sure sign you're missing something...
The problem is that this isn't your 3Com driver. An OpenGL driver has an incredible amount of high level code in it. In fact, from the moment you call a user-level API function, like glBegin() or glVertex() you're in NVIDIA's code. It's NVIDIA's libGL, NVIDIA's GLX module, and NVIDIA's kernel-driver. As a result, a lot of the optimizations that the drivers do could very easily be applicable to any OpenGL hardware. ATI could easily take NVIDIA's driver, and just replace the low-level code, while gaining the optimizations and OpenGL-completeness of the higher-level code.
A deep unwavering belief is a sure sign you're missing something...
To add on to this. NVIDIA's cards are really strange. According to some documentation at RivaTV (on sourceforge) the hardware has this object-oriented programming interface. Most other cards are much simpler, allowing you to set the properties of triangles in certain registers, with the card drawing the triangle when you've hit the last register.
A deep unwavering belief is a sure sign you're missing something...
I generally agree with this. Even back in the Riva 128 days, NVIDIA was a whole lot friendlier than any other chip company. Though, I really have to credit 3DFx. They are the only tech company on the planet whose manual writers actually have skills. Do yourself a favor and read the Voodoo3 spec. Its written like an actual document meant for people to read. Its got introductions that lay out the high-level scope of the section, supporting paragraphs that give details, examples for complication concepts, diagrams that are well referenced in the text, and concluding paragraphs that summerize the section. Absolutely wonderful! Compare this to an Intel doc, which consists of lists of registers and their semantics.
A deep unwavering belief is a sure sign you're missing something...
Nvidia has already released a driver for freebsd, where exactly have you been?
As best I can tell, not speaking Spanish, it's just a how-to to install the proprietary ATI drivers. I didn't see any comments about any particular games; of course, if I had I wouldn't have known whether they were good, bad, or indifferent.
Hey kids, there's only 5 days left 'til Yak Shaving Day!
I base my theory on two facts:
1) The GeForce2 hardware is identical to Quadro2 hardware, except for a single resistor. (which you can solder to "morph" your card into the other type - there are instructions for this on several websites)
2) The Quadro2 performs an order of magnitude faster than the GeForce2 on certain specific benchmarks, such as antialiased line drawing. These specific features are used in many CAD applications but not in any games.
Therefore, the driver must be crippling certain features when it thinks it's running on a GeForce 2. So NVIDIA can charge CAD users $600 for essentially the same hardware that a gamer buys for $200... (these prices were when the GeForce2 was first released, of course)
ATI didn't want pictures taken (maybe ATI does more hardware dev at Thornhill than NVidia does in Santa Clara?).
But giving access to the chip architects (more than a lunch meeting anyway) is cooler than a bunch of fuzzy pictures with Anand's thumb in front.
Did you write the drivers for your ATI card? Did you inspect them? Or did you trust someone else to do it? Why use an operating system written by other people? Why not write your own?
"Trust No One" is one of the stupidest reasons I've heard for supporting Open Source.
If NVidia wants to keep their source private, fine. Just so long as they accept the responsibility of supporting different systems. It would be nice if they did both, but writing closed source drivers is certainly better than *nothing*.
I bet you trusted someone else to write your sound card drivers too...
"Ignorance more frequently begets confidence than does knowledge"
- Charles Darwin
I trust no one corporation. Howsabout that?
Everytime you look at porn a devil gets their horns.
The problem with a unified driver that's proprietary is that despite what they say, older hardware does not get the same level of support as open source drivers for cards of the same age. Worse, since your architecture is unified, you can't open source the drivers for older cards without jeopardizing IP on new cards!
So, while I'm glad the unified driver works for you and your newish card, I had to ditch my TNT1 for an older Radeon because the unified drivers never supported my TNT1 on K6-III/VIA chipset very well (i.e., it crashed too much).
Cheers,
-l
who got a 64MB Radeon VE dirt cheap for his new flat panel. and yes, next year I hope to upgrade the mb/cpu. :)
Help cure AIDS, cancer, and more. Donate your unused computer time to worldcommunitygrid.org. Join Team Slashdot!
I guess you're too stubborn (or is that stupid?) to admit when you're wrong.
Dinivin
Besides Nvidia could very well release a BSD driver.
They have.
Now what kernel does the Mac OS X use?
The Mach kernel, which is NOT THE BSD KERNEL!
FreeBSD was the first OS the Mach kernel would run, which is why Apple based OS X on FreeBSD, but FreeBSD does not use the Mach kernel.
Under capitalism man exploits man. Under communism it's the other way around.
I used to blame nVidia drivers for my X crashes, too. Then I switched from KDE to WindowMaker (which I actually prefer) and no more X crashes.
Hmmm... I wonder whose code was buggy there...
Under capitalism man exploits man. Under communism it's the other way around.