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.'"

Re:A question

[pkphilip]

Though one may argue that having the source for a driver may result in the driver becoming more stable over time as a lot of people contribute changes/fixes to it, I feel that this may be overstated.

Consider Nvidia/ATI drivers on Windows or Mac OSX - these binary-only drivers are feature rich (are they more feature rich than their binary-only drivers on Linux?) and most users are quite happy. Bugs do occasionally show up, but they are normally fixed by Nvidia/ATI within a reasonable time frame.

However, I have noticed that these same manufacturers take forever to fix bugs which show up only on Linux.

That indicates to me that the reason that these binary drivers are not that stable on Linux is not because of the binary nature of these drivers but because the Linux user community matters less to Nvidia/ATI than the Windows user community.

And that is understandable - the number of windows users is roughly 93% http://http//en.wikipedia.org/wiki/Comparison_of_W indows_and_Linux/ [http] and Linux users probably account for about 3% to 4% for all computer users.

So it is understandable that a hardware manufacturer prioritizes bug fixes for their larger user base (windows) rather than for the Linux users.

Unless Linux gains the kind of market share which will force hardware manufacturers to take it seriously, we can expect less than stellar drivers and support from them.

Re:Video Acceleration Suppport

[sbryant]

I'm one of those people who are interested in better 2D acceleration and hardware MPEG-4/H264 acceleration. I'm a MythTV user. Nvidia's hardware does have support for such acceleration; they call it Purevideo. Unfortunately, the Linux drivers don't support it.

Well, the "nv" driver doesn't support it, but Nvidia's own driver does ("it" being the XvMC extension in Xorg).

Given that you're a MythTV user, here's a link to their Wiki, which contains a page about XvMC support [mythtv.org].

The new 7050PV [nvidia.com] chipset would be perfect for a high-definition ...

I've not tried it, but the 7300GS just came out a month ago, and I happened to be upgrading so I got one - for 40 Euros; it's passively cooled, and so far has been working really well in TwinView mode on my Linux box. I use Kaffeine (Xine-based) rather than MythTV, but it's very smooth. I'm sure MythTV would be the same.

I wish Intel would release a standalone video card.

BTW, I'm using a machine with an integrated Intel VGA at work. It's OK - has open source drivers, and 3D acceleration, but it uses system memory, only has one output (not DVI either), and doesn't always behave itself. Personally, I much prefer the 7300. Oh yeah - the Intel driver supports XvMC too.

-- Steve

Re:NDAs and Patents Suck Life.

[mikael]

What you say is confusing and has the smell of a well crafted lie. Can you set me straight so I can understand why Nvidia is unable to do like Intel and fully co-operate with the free software community?

Have a look at NVidia's OpenGL specifications web-page [nvidia.com]

Every extensions comes with an IP Status field. For example ARB_color_buffer_float has the following:

IP Status

        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.

        SGI will not grant the ARB royalty-free use of this IP for use in
        OpenGL, but will discuss licensing on RAND terms, on an individual
        basis with companies wishing to use this IP in the context of
        conformant OpenGL implementations. SGI does not plan to make any
        special exemption for open source implementations.

        Contact Doug Crisman at SGI Legal for the complete IP disclosure.

Re:Nouveau

[MoxFulder]

There is no way that an open-source GPU driver can ever achieve same quality (in terms of performance and compatibility at a given timeframe) as an actively developed proprietary driver. GPUs are very complex devices, and drivers make a huge difference on performance. In order to make a fast driver you really need to know how the target GPU works on a very low level.

This is the same tired, fallacious argument NVidia has been trotting out for years: "Developing GPU drivers is much too complex. Why don't you Linux kids go run along and play with your toys, and let us write the GPU drivers?"

Well, it's just wrong. I have used many reverse-engineered drivers for complex pieces of hardware, and nearly all of them work as well as or better than the original vendor's drivers. Most recently, I had the pleasure of trying out the Broadcom 43xx wireless driver, which was painstakingly reverse engineered [sipsolutions.net] in the face of an INCREDIBLY recalcitrant vendor that won't release a shred of documentation on their devices to open source developers, even under NDA! And, surprise, the card works better under Linux with the reverse-engineered driver than it does with the Windows driver, which seems to lose the signal quite often. The bcm43xx developer Michael Buesch has even got some evidence that certain parts of their code are implemented more efficiently and elegantly than the original driver.

Naturally, NVidia will not disclose this low level stuff about their GPUs to outsiders.

Why is this "natural"? Intel releases documentation on the low level stuff about their GPUs, Realtek goes out of its way to help the Linux community with Ethernet and wireless IC documentation, and Linksys has released the complete code for its Linux-based routers. If a business is based on continual innovation--rather than maintaining an entrenched monopoly--it is entirely possible to be successful and open at the same time.

You can compare GPU drivers to compilers. There is no way that open-source compiler (GCC) will ever produce as good code for new Intel's CPUs as Intel's own compiler (ICC) as long as ICC is actively developed.

Again, this is only true if Intel holds back some of the documentation necessary to make a good compiler. If they publish complete instruction set information, with accurate timing, cache, and pipeline data (which they have done, for the most part), then making a better compiler is "merely" a question of developer resources and talent.

I suspect that, to the extent that GCC code doesn't run quite as fast as ICC, it's because GCC has higher priorities for its code base, especially ensuring support for an incredible breadth of platforms [wikipedia.org]. GCC has essentially become the reference compiler for a lot of embedded development in particular. Even companies like Broadcom that won't lift a finger to help open source out rely on GCC to build the firmware for their own devices [broadcom.com].

Re:Nouveau

[the_greywolf]

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

No, I don't think NVIDIA are in the position to renegotiate a license that they don't have direct involvement in. I'll give a few examples from OpenGL extensions for why I think this way:

• ARB_occlusion_query -- HP claims the IP to this extension and licensed it to the ARB.
• ARB_fragment_program, ARB_fragment_program_shadow, ARB_vertex_program -- Microsoft claims ownership of the IP. What, precisely, they claim they own, I don't know. They're not likely to favor opening related source.
• ARB_texture_float, ARB_color_buffer_float, ARB_half_float_pixel -- SGI patent #6,650,327
• ARB_point_parameters, ARB_multisample -- "TBD," according to NVIDIA. It's probably not clear just who has the claim on it.
• EXT_texture_compression_s3tc -- S3 owns the IP on this specific compression scheme. It's the most popular and most widely-used compression format, so removing it might break a lot of games.

A number of other OpenGL extensions are NVIDIA proprietary, and most of the suits will probably hang on with their dying breath, if typical corporate behavior is any indication.

It'd be nice if they opened the driver, but half of the OpenGL ARB's members would probably sue their pants off if they tried.

It's bloody obvious what's really happening here..

[advocate_one]

Microsoft are the real owners of the SGI OpenGL patents and are blocking this entire show by insisting on those anti-GPL RAND terms for the licensing... not SGI...

http://www.theregister.co.uk/2002/01/16/sgi_transf ers_3d_graphics_patents/ [theregister.co.uk]

http://www.smithhopen.com/news_briefs_display.asp? ID=301 [smithhopen.com]

http://www.forrester.com/Research/LegacyIT/Excerpt /0,7208,28681,00.html [forrester.com]

that last is is a doozy... they want $99 for a one page article...

Microsoft has nVidia over the certification barrel... if they make the nv driver support 3D, then nVidia may find it very difficult to get their windows drivers certified... they're having enough problems at the moment...

SGI aren't the problem

[Anonymous Coward]

As far as SGI have said (very publicly) there is nothing that SGI have patents for that they don't have a problem being GPL'd