AMD Llano: Mobile discrete GPU killer?

[cbn]

Sure, but even on Windows, OpenGL is already problematic. Since so few games use OpenGL, driver developers don't spend a lot of time on optimizing and debugging the OpenGL drivers.
On the Mac the problem is even larger, so both Apple and AMD/nVidia have their work cut out, if they want to make the Mac competitive with Windows in terms of gaming.
I don't think that's where their priorities lie.

It seems to me gaming isn't Apple's priority either. What about graphical programs that could be used as "game-like" tools or simulations?

I would imagine in order for Open GL to take hold Apple would need to discover and/or create applications beyond what MS is already doing.

 

[IntelUser2000]

And how many of those people don't game on PCs because their PCs aren't up for the task?

I do have a friend who became like that. Used to play fair amount of games until he got a laptop which couldn't play much.

If the GPUs in the 2011 processors could really manage what they claim, I can see a ripple passing through the industry.

Hard to see it will plan as hyped though. Won't there be a chance for them to rise the prices beyond what current IGP solutions will afford?

 

[gorobei]

the first few generations of APU or cpu/gpu arent going to take anyone's breath away, but they don't really have to. They just have to be good enough to match IGP performance and offer some extra bells and whistles for the parallel compute side. As long as they are close, it is enough incentive for motherboard makers to get behind it since it is one less chip they have to buy.

Intel and AMD have already decided that this is the direction they are going. It is only a question of how long it takes the rest of the industry to utilize the new functionallity. As soon as programmers learn to use the combined out of order pipeline and the in order parallel pipeline together, the synergy should outweigh the demand for discrete/igp graphics in the mainstream market.

 

[Scali]

the first few generations of APU or cpu/gpu arent going to take anyone's breath away, but they don't really have to. They just have to be good enough to match IGP performance and offer some extra bells and whistles for the parallel compute side. As long as they are close, it is enough incentive for motherboard makers to get behind it since it is one less chip they have to buy.

Yea, I think the confusion was started because AMD tried to make people believe that their Fusion is going to deliver higher performance than ever (same 'native' lark as with Barcelona all over again).

Intel and AMD have already decided that this is the direction they are going. It is only a question of how long it takes the rest of the industry to utilize the new functionallity. As soon as programmers learn to use the combined out of order pipeline and the in order parallel pipeline together, the synergy should outweigh the demand for discrete/igp graphics in the mainstream market.

I'm not too sure about that. CPUs and GPUs operate pretty much completely independently of eachother, and as such a discrete videocard is a blessing, not a bottleneck. It's just 'fire and forget' for your parallel tasks, and it won't interfere with the rest of the CPU performance, as it doesn't eat your CPU's bandwidth and such.
So perhaps when programmers start to combine the forces of CPU and GPU, discrete cards only get in higher demand... who knows.

I think that would be a bit of a parallel with 3d accelerator cards. When everything was software-rendered, the videocard didn't have much of an impact on performance. So a lot of people gamed with low-end Trident ISA cards, instead of expensive localbus Windows accelerators and the like. For a game of Wolf3d, Doom or Quake, it didn't really matter.
But once 3d acceleration came around, the market for high-end cards suddenly exploded. You couldn't just use a budget videocard anymore, if you were serious about gaming.

 

[Medu]

Fusion is NOT about better graphics it's about a better CPU and a more consistence platform. AMD did not gamble $5bn on ATi to improve low end gaming systems they did it because they saw the potential that GPU's had at improving processing power.