From: alexd on
OK, first a bit of background; Nvidia make two lines of graphics cards, the
mass-market Gforce and the expensive CAD-oriented Quadro. ATI do something
similar. The two lines of cards are very similar, the substantive difference
being that the Quadro drivers provide support OpenGL under Windows; softmodding
cons the drivers into thinking that your Geforce is a Quadro.

http://www.techarp.com/showarticle.aspx?artno=539

Now, my question is, given that on Linux/X, hardware 3D acceleration == support
for OpenGL, is the concept of softmodding on Linux irrelevant? Could this also
explain Nvidia's reluctance to provide open drivers for their hardware [ie
fearing that someone might port them to Windows and erode their 'value add']?

--
<http://ale.cx/> (AIM:troffasky) (UnSoEsNpEaTm(a)ale.cx)
11:55:48 up 3:01, 3 users, load average: 0.25, 0.23, 0.20
"Stupid is a condition. Ignorance is a choice" -- Wiley Miller

From: Sheridan Hutchinson on
alexd wrote:
> The two lines of cards are very similar, the substantive difference
> being that the Quadro drivers provide support OpenGL under Windows; softmodding
> cons the drivers into thinking that your Geforce is a Quadro.

Could you clarify what you mean to say here; the standard Windows NVIDIA
Geforce drivers do very much support hardware accelerated OpenGL in
exactly the same way as the Linux drivers.

Is it possible that you've misconstrued the artcile?

> Could this also e xplain Nvidia's reluctance to provide open drivers for their hardware [ie
> fearing that someone might port them to Windows and erode their 'value add']?

An NVIDIA engineer who recently gave a long and detailed interview on
Phoronix stated in the article that the Geforce Windows and Linux
drivers share approximately 90% of the same code between them.

Given this finding, it is clear that we will not see a (decent) open
source driver from NVIDIA for Linux derived from their current codebase,
as this would effectively mean open sourcing their Windows driver as well.

--
Regards,
Sheridan Hutchinson
sheridan(a)shezza.org

From: alexd on
Meanwhile, at the uk.comp.os.linux Job Justification Hearings, Sheridan
Hutchinson chose the tried and tested strategy of:

> Could you clarify what you mean to say here; the standard Windows NVIDIA
> Geforce drivers do very much support hardware accelerated OpenGL in
> exactly the same way as the Linux drivers.
>
> Is it possible that you've misconstrued the artcile?

Absolutely. I assumed "hardware acceleration of professional 3D applications
like 3ds Max" was referring to OpenGL. For such a 'technical' subject, the
article is somewhat light on what the actual technologies involved are; all it
refers to is a list of applications and synthetic benchmarks.

--
<http://ale.cx/> (AIM:troffasky) (UnSoEsNpEaTm(a)ale.cx)
21:01:34 up 3 days, 13:15, 4 users, load average: 0.50, 0.32, 0.21
"Stupid is a condition. Ignorance is a choice" -- Wiley Miller

From: Tony Houghton on
On Wed, 21 Oct 2009 21:10:35 +0100
alexd <troffasky(a)hotmail.com> wrote:

> Meanwhile, at the uk.comp.os.linux Job Justification Hearings,
> Sheridan Hutchinson chose the tried and tested strategy of:
>
> > Could you clarify what you mean to say here; the standard Windows
> > NVIDIA Geforce drivers do very much support hardware accelerated
> > OpenGL in exactly the same way as the Linux drivers.
> >
> > Is it possible that you've misconstrued the artcile?
>
> Absolutely. I assumed "hardware acceleration of professional 3D
> applications like 3ds Max" was referring to OpenGL. For such a
> 'technical' subject, the article is somewhat light on what the actual
> technologies involved are; all it refers to is a list of applications
> and synthetic benchmarks.

I had a look at <http://en.wikipedia.org/wiki/Nvidia_quadro>. The main
difference seems to be that GeForce is optimised for speed at the
expense of accuracy and Quadro vice versa. AIUI the differences aren't
just in the driver, but in the firmware too, and there are sometimes
other hardware differences.

--
TH * http://www.realh.co.uk