NVIDIA's Andy Ritger On Linux Drivers

tykev writes "The Director of Unix Software at NVIDIA talks about Linux drivers, planned features, development cycle, and the open source Nouveau driver. (The interview is in English but all the comments are in Czech.) Quoting: 'NVIDIA's stance is to neither help nor hinder Nouveau. We are committed to supporting Linux through a) an open source 2d "nv" X driver which NVIDIA engineers actively maintain and improve, and b) our fully featured proprietary Linux driver which leverages common code with the other platforms that NVIDIA supports.'"

Nouveau

[nxsty]

NVIDIA's stance is to neither help nor hinder Nouveau. We are committed to supporting Linux through a) an open source 2d "nv" X driver which NVIDIA engineers actively maintain and improve, and b) our fully featured proprietary Linux driver which leverages common code with the other platforms that NVIDIA supports.

But what will they do when nouveau is complete, and replaces the nv driver? Will they stop commiting to xorg?

NVIDIA driver team members use Debian?

[Karellen]

From TFA:

"Across the NVIDIA Linux Graphics Driver team, everyone has their own favorite Linux distribution as their primary desktop: Debian, [...]"

Interesting, given that Debian can't ship their driver.

Oh, I know that none of the driver team will be using a distro-bundled version of the driver anyway, but still...

Re:Nouveau

[Andy Dodd]

Yeah. I interpret their "help but not hinder" comment to mean:

"Some of our license agreements prevent us from contributing 3D support to open source drivers, we like Nouveau but can't legally contribute to it."

One wonders what happens if Nouveau becomes the official xorg driver. Perhaps NV will be able to legally contribute to the 2D aspects of the driver. The binary drivers will likely continue to exist as long as they are encumbered by technologies licensed from other vendors which NV has no direct control over (such as S3TC support).

I'll still switch..

[pjr.cc]

ATI said they'll go OSS with their drivers and if they do i'll switch away from nvidia and be happy to do so given that article.

Re:A question

[mrcgran]

One big problem in not releasing the source code is that they actually are not making our computer experience better: their drivers have bugs, and we will be locked to whatever features and bugs THEY want to make available to us -- so, basically we become hostages of their will, they can do whatever they want, because whithout THEIR driver, your nvidia card isn't worth its weight.

In the future, when new versions and extensions of OpenGL are released, we won't have any guarantee that they will properly update the drivers. So, you'll probably won't be able to use their proprietary drivers in 5 years for new applications (shining new wobbling effects), because these apps will need new extensions, but the driver for your specific nvidia card is arbitrarily not supported anymore by them (they want to force you to throw the old one away). Too bad for you.

On the other side, if we have access to the source code (or at least the hardware specification), we don't even need nvidia's help: we can do the updates/bug-squatting ourselves, much better than a small team at nvidia. This is something that these companies don't get: the whole world is willing to write their drivers for free and maintain them to the end of times, but they refuse the consumer this right (or maybe they get, they just want you to throw away your old card and buy a new one). We don't want a huge amount of work from them, quite the contrary! It's *way* cheaper for them to release an open-source driver: it costs nada/zero, we can build one with the bare bones of a reasonable hardware specification, a little pdf file -- how much does it cost to post a pdf file on the Internet?

There's no RMS ideology in that, only the absolute minimum someone would expect in terms of support for something you bought. Nowadays, the choice is clear: go Intel X3000/X3500 [wikipedia.org], which supports open source and you can be sure will always be up-to-date. Ignore nvidia and ati, until at least one day (I hope so) nouveau arrives.

nVidia PS3/Linux Driver?

[Doc Ruby]

How about a driver for Linux on the PlayStation 3 [psubuntu.com], which would let the PS3 RSX chip actually work for Linux apps?

Right now PS3 Linux runs all display processing on the PPC core on the Cell, which needs to do a lot of other processing to keep the complex Cell going. Meanwhile there's an RSX chip that runs at 1.8TeraFLOPS, dwarfing even the Cell's 0.2TFLOPS. But Sony's Hypervisor virtualization layer that runs Linux hides the RSX from Linux. However, the RSX is exposed in some API, otherwise PS3 Linux wouldn't display on the HDMI port out of the PS3, and sound probably wouldn't work (probably also running on the RSX somehow).

Sony doesn't want the RSX exposed to Linux apps, because then Linux apps could compete with Sony-licensed games (without paying Sony the royalty that even subsidizes over 25% of the PS3 purchase price). But can't nVidia release a driver, or some kind of specs, that expose a 2D API for running X desktops? Sony's money all comes from 3D games.

Or maybe someone else has a way.

Re:Nouveau

[b1ufox]

The problems as i look at it is loosely like this. Nouveau being a reverse engineered driver misses to harness complex Nvidia specific GPU features.Yes, mod me down if you feel like but that is the truth, a reverse engineered software for a complex GPU is not simple.And now please don't give me example of Samba and et all, i know it. The way hardware evolves in today's hardware industry Nouveau will find it extremely difficult to sustain the pace to match it.

Moreover reverse engineering being a difficult task, it will take time.

Here is the alternative IMHO, AMD open sources ATI drivers. ATI 3D driver gets into mainline kernel. Nvidia will *have* to open source their driver, which can be possibly merged with the nouveau project to make sure it is constantly up to date with the Nvidia's cutting edge technology advances.

But thats just optimistic me. :-)

Re:Nouveau

[babbling]

Do you really think Nvidia couldn't renegotiate license agreements, work around patent problems, or at least release a data sheet for their cards?

It's all a matter of how desirable it is for them.

If Nouveau became the official xorg driver and was decent enough that people wouldn't rush to replace it, Nvidia most likely find a way to contribute to it to ensure that Nvidia hardware has a good quality driver on Linux.

Linux-certified hardware

[athloi]

Windows certifies hardware, and Apple makes it clear what they support. Could it be useful for an agency of Linux developers to certify hardware that is open (standards released so drivers can be written) and well-designed enough to support the rigors of a "UNIX-like" OS?

I do not know the answer to this one. My inner four-year-old anarchist is leery of certification in anything, but even something as simple as a list of supported hardware like BSD does, with the requirement that its standards be open so drivers can be developed, might help companies market to Linux users (1 in 10 users, by my estimate) and help Linux users get their market share behind a few quality products so they can stand up and be counted.

Just an idea. Feel free to mod -1, this guy's an idealistic moron.

Re:Video Acceleration Suppport

[kimvette]

Instead, I think the way to go is with Intel integrated GPUs. They don't have MPEG acceleration yet, but they are working on APIs and drivers to greatly improve this (the hardware has very good video support). I'm hoping Intel GPUs quickly become the far and away best option for Myth users.. let open source prevail.

Intel open source drivers are fine and all but before I abandon Nvidia for Intel I need:

  - Dual dual-link DVI ports (the nvidia cards we buy have these)
  - HDTV & S-video ports
  - dedicated video ram
  - in a discrete card (are there add-in video cards based on Intel GPUs?) - if I need to upgrade the video, or replace a bad video card, why the hell should I have to upgrade the board (and processor, and CPU) all at the same time?

I, for one, am hoping that AMD forces ATI to open up their drivers. THAT will create a ripple effect through the entire industry. Nvidia will have no choice but to follow suit.

Re:Nouveau

[jonwil]

I wonder if it would be possible to produce a driver that didnt use the patented features (and as such presumably didnt support that relies on the patent) but still implemented the rest of the functionality of the GPU. For example, if nvidia has licensed the S3TC patent, they could remove the S3TC code from the "open source" driver.

Or they could go half way by taking all the code that they don't own and all the code that relies on 3rd party patents where the patent holder has forbidden them from releasing source and put those into the "binary blobs" (like they have now) and make the rest of the code (which they own the copyright to and which is not covered by whatever 3rd party patents they have licensed) available (in much the same way at least one vendor of WiFi cards has put only the bare minimum code that has to remain closed for FCC certification into a binary blob and made the rest of the code open source)

Re:Nouveau

[WilliamSChips]

The problem with that is that AMD is bound by the same patent crap nVidia is on a lot of their 3D features and neither one can actually open their drivers because it wouldn't be worth the patent minefield to navigate around. But if Nouveau advances far enough to supplant nv, then nVidia will probably send a few engineers to contribute to Nouveau code like they currently do with nv and forcedeth(nForce network driver for Linux)

Re:Nouveau

[ppanon]

My understanding is that GCC doesn't perform as well as ICC because ICC uses a number of patented algorithms in its optimizations which GCC cannot yet legally use. Not sure if somebody in Europe might be able to fork a version of GCC that uses those algorithms for use outside the US. The problem of course is that many of the maintainers of GCC work for Cygnus Software (now part of Redhat?) and are based in the US.

OK, that Sucks Life.

[twitter]

Thanks for the link, it clears up a lot. The site itself is a pain, so it would probably be easier to Google site:developer.nvidia.com for "IP Status".

As for the example you give, Holy Shit!

SGI owns US Patent #6,650,327, issued November 18, 2003. SGI believes this patent contains necessary IP for graphics systems implementing floating point (FP) rasterization and FP framebuffer capabilities.

A patent on floating point raterization and framebuffers? Is that what I think it is? Yes [wikipedia.org] it is [wikipedia.org]. I can not think of anything more obvious in high quality imaging than representing the image as a floating point matrix. It may be true that there are still "fat line" patents out there [forbes.com].

Kudos to Nvidia for shining a small light on this insanity. Knowing the problem is always the first step. It would be nicer if they would put patent and other encumbering as symbos on the reference page and a link to the actual patent in the description page.

Re:Nouveau

[MoxFulder]

I don't see your point. You talk about ethernet drivers which are ususally low complexity drivers and use this as an argument to support your 3d driver claims.

Full-featured 802.11g wireless drivers are pretty darn complex. I don't believe there is anything particularly "exceptional" about 3D drivers, so I see the comparison as perfectly valid. For years we've seen proprietary software makers concede that "open source can do X, but Y is too complicated." I have no reason to believe that 3D drivers are any different.

People have *talked* about developing 3D drivers for modern GPUs for *years* and go figure they don't exist or are even anywhere near existing today. Its a horizontal market and there is a lot of demand for 3D so I don't see the excuse. I'll believe it can be done when someone actually does it.

Well, the Intel GMA drivers already exist, having being released by Intel as fully open-source. And the open source community is (surprise!) working on improving them.

And the Nouveau Project seems to be making quite a lot of progress [freedesktop.org] on NVidia drivers.

As far as Intel... lets see Intel sells general purpose CPUs that need to interoperate within a wide range of operating environments. Intel has to release the specs in order for this to be achivable to the mutual benefit of all parties.

NV and AMD are interested only in winning their bloody GPU war and have no interoperability requirements or constraints. Giving away their secrets/work/whatever you want to call it at this level is of no advantage to these companies and I think this is all pretty obvious. Why is this so hard for some to understand?

I don't think this is quite accurate. As long as it's only NV vs. AMD, they may not have much reason to release anything. However, this creates an opening for a third player, who will start out behind but gain a competitive advantage by releasing open-source driver.

And at this point, Intel is clearly gunning for that spot. They already have an extensive line of integrated graphics, and they'll be making discrete graphics [slashdot.org] chips soon. If Intel follows through and releases open-source drivers for these new cards, I expect that open-source developers will flock to them in droves and rapidly improve the drivers. This will put more pressure on NV and AMD.

Well, time will tell which of us is right :-)

Re:SGI aren't the problem

[ichigo 2.0]

They could always change their mind. The GPLv3 would help in this case only if SGI themselves distributed/contributed code that covered the patent in question. If they are really serious about not minding GPL code that use those patents, maybe they will release something under v3 themselves? Just saying something publicly doesn't make it legally binding.

Re:Nouveau

[Andy Dodd]

Regarding the last one (S3TC) - removing it WILL break a lot of games.

Unreal Tournament's need for the last one marks the beginning of ATI providing proprietary drivers and partial specs for their cards rather than full specs. There was a huge to-do about Unreal Tournament 2003 only working on NVidia cards in Linux back when UT2K3 was released. Epic's response was basically "we use this feature, it's implemented in all Windows drivers and the NVidia 3D drivers, we cannot work around the lack of this feature in the ATI drivers."

Shortly after, the first binary ATI drivers were released, with the main difference between them and the open-source drivers being S3TC. They have diverged since then, as clearly a number of other encumbered extensions have gone mainstream after looking at your list.