in the pc space, their fusion chips I find to be highly fascinating, but ultimately not really competitive vs a traditional intel cpu + discrete GPU.
Yet they power the two big consoles of the next-gen.
The life cycle of a PC SKU is about two years for AMD (launch + simple re spin for AMD, like Trinity and Richland), and they are trying to move their small core line to a yearly cadence, so whatever the R&D money they spent they must recover it in this short time frame.The consoles on the other hand, are warranted to get orders until 7 years down the road, maybe even more, so AMD can spread the investment recover for a lot longer.
Another factor is the supply chain. Sony doesn't need to make money on the PS4, they need to cover the costs. The equivalent of Sony in the PC supply chain are the OEMs, and they *need* to make money on each PC they sell, so they should have higher margins on a PC than Sony in a PS4. Don't forget to add to that higher price the top notch cooling, higher end MB, GDDR5 memory and everything else this super APU would need.
The result of this is that AMD can swallow a 10% margins on console but can't the sub-40% margin of the Bulldozer line, so if AMD were to launch a console APU on the PC market, it would be far more expensive than their current chips, and that for a chip that would perform poorly in most workload outside games extremely optimized for it.
Just to put things in perspective, AMD ASP before APU was below $40, and the console APU themselves, which are slightly above the $100 mark. That APU would have a ASP at AMD's door of at least $150 if AMD were to recoup their money. I can't fathom this processor costing less than $300 + taxes + the more expensive platform, best case, to the end user.
In the end, I don't think there's a solid business case here.