Originally posted by Michael
View Post
Announcement
Collapse
No announcement yet.
Open-Source AMD Fusion Driver For Ontario Released
Collapse
X
-
Originally posted by crispy View PostBy the way, this AMD open source approach makes a lot of sense from a business perspective also. Since better open source support means better android support is AMD moving into this market?
And please correct me if im wrong
Comment
-
AMD did, and what you find related to those AMD assets that they sold off to qualcomm, are the "adreno" parts associated with snapdragon SoC's.
Now here's the crazy thing that I'm seeing....
The UVD parts that AMD continues to use are evolved from the ATI-AMD Xilleon, which was sold/shared to.... BROADCOM, and no doubt is the basis of the 97001x video decoders, FOR WHICH THERE ARE OPEN SOURCE DRIVERS.
So Xilleon is the basis of BOTH UVD as well as CRYSTALHD.
Just how much of crystalhd is applicable to uvd?
It would be real nice to see open source drivers working the UVD.
Comment
-
Nice to see
This code dump happened much quicker than i thought. If power management works reasonably on netbooks equipped with such a chip, i'd happily ditch my current atom netbook for a fusion-based one.
Seemingly, the past few years of playing catch-up and building a platform for decent OSS drivers (with the rest of the OSS driver community, of course) are paying off... well done!
This also fits well with the MeeGo joining; my guess would be that AMD wants to get a piece of the most trendy cake available at the moment - the tablet market.
As for UVD, I'd rather want a decent VA-API implmentation over shaders in gallium than hackish UVD support that can get AMD into legal trouble with the content mafia (leading most definitely to trouble for the amd-employed OSS driver coders and their rather OSS-friendly strategy). If we already have gallium sort-of-working, why not use it?
Comment
-
Originally posted by wirrbeltier View PostAs for UVD, I'd rather want a decent VA-API implmentation over shaders in gallium than hackish UVD support
that can get AMD into legal trouble with the content mafia
If we already have gallium sort-of-working, why not use it?
1) power consumption,
2) performance.
Gallium can *help* the power consumption and performance problems only, can't come even close to curing it. How would you like to have a fusion tablet that can only play 10 minutes of 1080P, CHOPPY, before running out of battery?
Comment
-
Originally posted by droidhacker View Post[..]
Broadcom has proven that it can be done safely. Content mafia be damned.
[...]
Comment
-
Originally posted by glisse View PostBroadcom isn't in the same situation they aren't tie with DRM content protection agreement, broadcom is doing wifi stuff mostly, nothings such as GPU. The content mafia is all about picture & music, they don't care about wifi.
Broadcom chips are even baked into TV's, DVR's, and BD players ***FOR*** image decoding and processing.
They most definitely ARE involved (sucked in to) DRM content protection agreements.
Of course, if you'd read this thread, you would know that already, since this is now the SECOND time I've explained it.
Comment
-
Originally posted by glisse View PostBroadcom isn't in the same situation they aren't tie with DRM content protection agreement, broadcom is doing wifi stuff mostly, nothings such as GPU. The content mafia is all about picture & music, they don't care about wifi.
Comment
-
Originally posted by droidhacker View PostBCM970010, BCM970012, BCM970015 -- based on Xilleon, which they got from AMD, which is also the predecessor to UVD. SAME STUFF, EXACTLY THE SAME ISSUES.
Broadcom chips are even baked into TV's, DVR's, and BD players ***FOR*** image decoding and processing.
They most definitely ARE involved (sucked in to) DRM content protection agreements.
Of course, if you'd read this thread, you would know that already, since this is now the SECOND time I've explained it.
And where can i get opensource driver and documentation of their chipset ?
Comment
Comment