edit: nevermind

Lol. A 6670 (which is a 40nm chip) uses less than a third power than a 7970. A modified 6670 in 32nm wll use less than a fourth the power of a 7970.

And in an intergrated chip you want as low power and effeciant use of space possible meaning a more modern achitecure is a must. And would make bassing it off an older desktop GPU even less likely as you would basically need to change everything about it to get it to work well on a SoC. AMDs Trinity GPU core would be the best bet bassed on the 7000 series it is already designed for a SoC and has a groundbreaking CPU/GPU linking system that promises much better performance and would be perfect for SoC bassed console, again that makes the IGN rumour look suspect. 

Fu&*! if this is true. I will just buy a mid-range pc next year instead.

not really 45nm to 32nm only offers a ~30% reduction in TDP, 40nm to 32nm would certainly drop the TDP below a 7970 but the performance per watt would still be nowhere close.

I'm liking the HW discussion here, you guys fill me with hope for this community. xd

OT:

Any rumblings on other HW that will be in the dev kits?

At least hoping for:

4 GB RAM at a decent speed
IBM (Which sounds likely if IBM is actually producing these combo chips) 6-7 core proccesor @ 3.2+ GHZ.

Off topic:

Didn't they test the CPU/GPU combo in the 360S btw?

Read somewhere it actually gave them better performance than on the older models which had the CPU and the GPU ''seperate''.

@Mr. Khan said....

"Sounds reasonable to me. It's more likely in my mind that Microsoft and Sony aren't going to take as big a leap with their next consoles as they did with the 360 and PS3. One easily forgets that PS360 were easily 3 years too soon, as evidenced by their unprecedented sales patterns."

How does one 'easily' forget that the PS360 were released 3 years too soon??... lol.
I'm pretty sure, well absolutely sure actually, that neither console was released early. I think they both came to retail at the exact time they're respective producers intended.
I think what you're easily forgetting, or not recognizing all together, is that the PS3 and 360 are drastically different video game consoles than they're predecessors, incorporating so many different types of entertainment into single, convergent-type machines, that basing expected lifetime market performance on the historical sales curve of prior consoles is completely inaccurate, if only because those previous game-decks were just that, simple gaming consoles.
These ain't your daddy's video game systems so to speak.
I think what you term as '3 years too soon' is in reality a testament to the amazingly, long-term design approach that both Microsoft and Sony employed when building their consoles, ensuring themselves a stable gaming platform that would be powerful enough to at least keep pace with the flux of technology as the general public sees it, and adaptable enough to fool those people who pay closer attention.
If you look at the 360 today.....
'one easily forgets this is the same console that launched with an entirely different user interface, a comparably weak, if almost non-existent, menu of alternative entertainment options, and a game catalogue that included the likes of 'King Kong' and 'Perfect Dark Zero'
Those unprecedented sales patterns you referenced are a result of the intuition, hard work, and opposition to convention that both companies have demonstrated in growing their consoles. The network of games and entertainment options offered is like absolutely nothing before, so why would the sales patterns of either follow historical record?

I don't believe rumors. I read them and I consider them, but I don't believe them. I believe official announcements. Until that happens I don't have any expectations.

Weak, but not surprising consider how much room and power would be required to go for a huge jump.

the PS3 was even delayed because of the blu diode shortage and low CELL yeilds...