It would be interesting but the fact it's Intel would make it not a good choice. Their chipsets are a joke and I can't see an APU-like unit running on anything other than laptops. Next gen is pretty under powered to begin with.

Price quality ratio. I think that's the reason.
We all know how nvidia can be expensive ~

Actually if you've being keeping a close eye the iris pro 5200 is a fairly good solution since it beats the A10 5800K while using only half the TDP of the A10 5800K. Broadwell will probably finally be able to match amd kaveri. 

With my Graphics Programmer cap on

Tessellation is way down the list of useful GFX tech

Last year nVidia chips used too much power. The 7790 which is I think is closest to what's in the consoles, was at least 1.5X more power efficient than equivalent nVidia GPUs, until this year.

PC Fangirl detected.

~ Mod Edit ~

This user was warned by TruckOSaurus

I was pretty sure that kepler was more efficient perf/watt wise in terms of gaming performance ...

Bump ...

I disagree.
Geometry is going to play a massive role this coming generation.

nVidia stated that the jump from the NV30 (GeForce FX 5800... Aka, Geforce 5.) to the GT200 (GeForce GTX 280... Aka Geforce 10.), the geometry performance of NVIDIA’s hardware only increases roughly by 3x in performance, that is in stark contrast to the shader performance of their cards increasing by over 150x.

To put that into perspective, that's just slightly more than the generational jump between the origional Xbox and Xbox 360.

However, nVidia also states that comparing the GT200 (Geforce 10), GF100 (Geforce 11) geometry performance has increased by a relatively wopping 8x.
And those are "old" architectures compared to Graphics Core Next, which should have an edge over the Geforce 400 series in Geometry.

Tessellators can provde several advantages:
1) Simpler models, then use Tessellation to give them detail, potentially increasing performance.
2) Massively more detailed models.

However, by generations end, I fully expect horrible "flat" surfaces will be a thing of the past.
The PC can take advantage of that fact as you can "adjust" the Tessellation factors to provide increases in image quality, whilst consoles will be stuck with whatever developers hand to them.


Tessellation itself as a technique has been used fairly sparingly over almost a decade and a half, so it's not a "new" technique, but it is "new" for it to be supported in all hardware and API's and devices (Aka. Consoles).
So I am indeed excited for what it will bring to image quality and it should not be understated.

AMD's design is easily less elegant than nVidia's, I'm not sure if AMD is still using the GDDR5 memory to store the Tessellation data with Graphics Core Next like they did with VLIW4, but that would bring with it, it's own performance penalty. (Higher latency, lower bandwidth.)
If anything, the Xbox One could have the advantage over the Playstation 4 in that regard thanks to the eSRAM, but, we will only know for sure in time, it certainly won't make up for the shader, render output, texture mapping unit deficiencies though.

Essentially, AMD needs to do allot of work to "Load Balance" the geometry, which comes with it... A bunch of bottlenecks, nVidia's Polymoph is much more scalable.

This is one of the biggest fail threads I have ever seen. In most games you can turn down (or off) the tessellation and not tell the difference without looking really hard. AMD cards are classically better at larger textures and AA, which is far more noticeable.

On top of that AMD was the ONLY option in terms of pricing.
1) Nvidia's CPU's are awful.
2) Intel's graphics are awful
3) AMD has good CPU's and great graphics.

P.S. The 280X loses to the 270X in your "benchmark." That makes absolutily no sense so good job finding the most non-nonsensical benchmark I have ever seen.