The UVD/UVD2 thread.

[eugrus]

Last I heard UVD was not going to be available for fglrx. They would instead push XvBA as a replacement.

But still... there are some plans for Linux/AMD video acceleration? Any dates?

I'm gonna buy a new AGP video card. Unfortunattely there are no VDPAU capable chips installed on AGP cards.

So, I thought Radeon HD would be the best buy for video acceleration hopes?

[Kjella]

But still... there are some plans for Linux/AMD video acceleration? Any dates?

I'm gonna buy a new AGP video card. Unfortunattely there are no VDPAU capable chips installed on AGP cards.

So, I thought Radeon HD would be the best buy for video acceleration hopes?

Long story short, UVD/UVD2 is a legal quagmire. First they'll finish what they're doing on 3D acceleration, then power savings, then try gettin git through legal and get it working... Earliest 2011 if you're lucky, if it ever happens is my guess.

The alternative is doing a generic implementation on shaders using Gallium3D. Some attempts have been made, certain things can be accelerated and others not, Nobody's really sure how much processor power you'll still need, but we know it's much more than for fixed function hardware, noone knows how many shaders it'll take.

Those who can do it are probably also mostly busy working on 3D acceleration of OpenGL, OpenCL and whatnot else. In short I haven't heard of any that are really working on it right now. so it's all up in the air. I don't think it'll happen very soon.

Not what you want to hear I guess, but your best bet right now is getting something new and nVidia. Personally I'm looking very closely at the recently launched Acer AspireRevo, tiny little nettop with 9400M and 1080p decode...

[cyrus_mc]

Long story short, UVD/UVD2 is a legal quagmire. First they'll finish what they're doing on 3D acceleration, then power savings, then try gettin git through legal and get it working... Earliest 2011 if you're lucky, if it ever happens is my guess.

The alternative is doing a generic implementation on shaders using Gallium3D. Some attempts have been made, certain things can be accelerated and others not, Nobody's really sure how much processor power you'll still need, but we know it's much more than for fixed function hardware, noone knows how many shaders it'll take.

Those who can do it are probably also mostly busy working on 3D acceleration of OpenGL, OpenCL and whatnot else. In short I haven't heard of any that are really working on it right now. so it's all up in the air. I don't think it'll happen very soon.

Not what you want to hear I guess, but your best bet right now is getting something new and nVidia. Personally I'm looking very closely at the recently launched Acer AspireRevo, tiny little nettop with 9400M and 1080p decode...

2011? I honestly can't see it being that long. By that time, most of the cards that have UVD support now will be obsolete.

[bridgman]

Kjella is talking about the open drivers, not fglrx -- that's where the legal issues are tricky because of DRM considerations.

I expect you'll see UVD support on fglrx before you see it on the open drivers.

[Kano]

I guess before you see that in fglrx then cheap dual cores reach 3 ghz limit that do not need any accelleration

[Antharian]

Kjella is talking about the open drivers, not fglrx -- that's where the legal issues are tricky because of DRM considerations.

I expect you'll see UVD support on fglrx before you see it on the open drivers.

Are there any road maps for UVD support on fglrx? I've been hanging on to the hope that ATI will eventually have support for High Definition playback, but my hopes are fading not to mention the WAF (Wife Acceptance Factor) is at its end.

Thanks,

-a

[Vighy]

Are there any road maps for UVD support on fglrx? I've been hanging on to the hope that ATI will eventually have support for High Definition playback, but my hopes are fading not to mention the WAF (Wife Acceptance Factor) is at its end.

The WAF needs to be taken seriously into consideration. Otherwise it's f**king hell!

But what makes me wonder what's happening, is the fact that they already claim to detect UVD2 (on UVD hardware too) since months ago but since then nothing happened.

Should I consider Nuke Dukem Forever much more imminent than UVD support!?!?

[bridgman]

Probably not

[bridgman]

The problem is that our requirement is not to "protect BluRay until someone cracks it", it's to "provide a ruobust environment where protected video content can be processed and displayed securely". Hacking BluRay encryption doesn't change anything, unfortunately.

[MartjeB]

Don't blame ATi, blame Hollywood..