NVIDIA PR Responds To Torvalds' Harsh Words

[Hirager]

well, it's gets better or worse: http://www.brightsideofnews.com/news...-and-nres.aspx

Wow, just wow - I am speechless...

[Licaon]

well, it's gets better or worse: http://www.brightsideofnews.com/news...-and-nres.aspx

[GreatEmerald]

Na the farthest back I can see them dropping support for is back to the FX series (GF 5). G92 era cards are still out there in abundance and have compute capabilities and their current driver easily handles GF6+ cards.

...except that it has already been dropped. And a while ago. First they dropped everything up to the FX series, and then they dropped the FX series itself, at least from what I can see on the openSUSE wiki. And sure, there are plenty of older cards on the market - but you didn't see that stopping AMD, did you? And I'd imagine that there were still FX cards being sold when they dropped support for that.

[figuerom]

This has been interesting to learn about.

I was unaware of the licensing issues between the binary Blobs and the Linux kernel.
I was also not aware that Nvidia might not be supporting Wayland. Or at least very slow to.

I'm not opposed to using blobs, but I understand the importance of even having the mere open-source drivers exist now (which I didn't before).
I'll probably avoid Nvidia now only on the basis of that future support might be terrible. Which could be an over-reaction on my part.

Right now I'm about to buy a new computer (on a budget and buy a graphics card later... maybe). I was really on the ropes between buying a new AMD Trinity APU + Ati or Ivy Bridge + Nvidia.
But after reading all these comments and links I've decided on getting an AMD Trinity APU.
AMD could be slow to support Wayland in the Blob; I'll have an Ok open source driver to fall back on. I'll also have the chance to use Crossfire if/when that gets put in when I'm not happy with the speed of games. I was going to get Intel Ivy bridge but I want to do some decent gaming without discrete graphics card even though Mesa is pretty solid for Intels IGP. I think AMD will improve faster in its Blobs and open-source though I could day overly optimistic about AMD also.

[deanjo]

Yes, they are a bit overdue with dropping support for their cards, aren't they? I expect that by next year, all cards older than the 200 series will be dropped from active support. They already dropped support twice for older hardware, after all.

Na the farthest back I can see them dropping support for is back to the FX series (GF 5). G92 era cards are still out there in abundance and have compute capabilities and their current driver easily handles GF6+ cards.

[kotakotakota]

Yes, they are a bit overdue with dropping support for their cards, aren't they? I expect that by next year, all cards older than the 200 series will be dropped from active support. They already dropped support twice for older hardware, after all.

I hope not... My main system runs a 200 series, and all my others run hardware that is older than that....

[GreatEmerald]

Another one of my major concerns is for old hardware support. All hardware gets old, and I can't afford to upgrade whenever something new comes out, so I find it to be very important. While it's true that the open source Radeon drivers do a fine job with supporting archaic hardware, the same is not true for AMD's proprietary alternative, fglrx. Nvidia however has been fairly good about maintaining support for older hardware. While this is not related at all to open source vs closed source, I think it's a definite plus of Nvidia's closed source driver, and has kept me a happy customer.

Yes, they are a bit overdue with dropping support for their cards, aren't they? I expect that by next year, all cards older than the 200 series will be dropped from active support. They already dropped support twice for older hardware, after all.

[kotakotakota]

Another one of my major concerns is for old hardware support. All hardware gets old, and I can't afford to upgrade whenever something new comes out, so I find it to be very important. While it's true that the open source Radeon drivers do a fine job with supporting archaic hardware, the same is not true for AMD's proprietary alternative, fglrx. Nvidia however has been fairly good about maintaining support for older hardware. While this is not related at all to open source vs closed source, I think it's a definite plus of Nvidia's closed source driver, and has kept me a happy customer.

[johnc]

Ohh, and I'll buy the best GPU for the job, depending on what my needs are for the particular system I'm building.

I am not beholden to NVIDIA or any other GPU vendor. I balance the needs with costs and will buy the best one for me.

[johnc]

Like what? Not for closed driver reasons? Let's hear what the Android crowd has to say about that:
http://forums.developer.nvidia.com/d...ra-2-3-etc-/p1

The chart you posted... the GPU market distribution has nothing to do with open vs closed drivers, as 99% of that market doesn't even have access to open drivers.

And then what about John Carmack, who cries for low level, direct access to graphics cards on PC's? He really raged (pun intended) with the last id's release about the abstractions of graphics card drivers slowing down performance to such an extend that he favors Intel graphics because Intel lets him actually access the low level memmory management and memmory blitting, without having the graphics driver needing to load an entire tile of texture memmory before a single pixel can be read out (see also Battlefiel 3 PC horrors):

http://www.pcper.com/reviews/Editorial/John-Carmack-Interview-GPU-Race-Intel-Graphics-Ray-Tracing-Voxels-and-more/Intervi

And now that AMD allows direct access with Fusion and documentation, where does that leave nVidia?

Face it man, nVidia realy does SUCK. PS: They are the new 3dfx. I hope you will not clamp onto a GeForce like the 3dfx fans clamp onto their Voodoo's, and just move on.

He is referring to the advantages of IGPs.

What advice would you give to NVIDIA, who have no chance of creating an IGP in the x86 space?

You are not seriously suggesting that if NVIDIA offered specs and documentation that the open source community could provide a better-performing driver than the one released by NVIDIA?