I think the catch from moving away from a Cell architecture would be the difficulty in trying to achieve backwards compatibility. If Nintendo and Microsoft both have BC at launch, it would definitely hurt Sony if they do not have it.

I'm sure that Sony is weighing the costs of R&D to improve the Cell, versus investing in known technology.

I just hope that the PS4 doesn't turn into a glorified PC.

I disagree with a few of your assessments. I deleted some text to make it easier to read.

The problem with using an X86 architecture is the fact that the chips themselves give much higher margins as desktop/server parts. What incentive does even AMD have to sell them X86 chips for the typical $50-60 range when they can get twice that selling essentially the same chip to OEM computer manufacturers? If they sell them a cheap chip then the performance may not be higher than what they have at present. With AMD vs Microsoft negotiations, Microsoft can sweeten the deal by implementing AMD specific extensions in Windows and raise the potential selling price of all AMD chips whereas Sony does not have the same negotiating position. 

Also if they do move towards a more GPU compute orientated architecture then some derivative of Fermi would be their best bet as Nvidia has the best compute orientated architecture in the GPU business on both the hardware and software development arenas. It would also further complicate their ability to get backwards compatibility for their PS3 games and complicate things even further (see Microsoft for Xbox 1 BC).

My question is, Will it be cheaper than a PS3 at launch?

PowerVR shoots for mobile environments these days with their design, not so much on raw power that you plug into a wall, Nvidia is stepping into that market now as well, it will be interesting.

you guys who say Cell is a mistake, I doubt that's really the case personally, its the high entry price tag and being one year late with shitty online when it first came out that really hurt the PS3 so devs didn't go into it as hard.

@Squillam

AMD is far below fab capacity. If someone offered to buy vast quantities of chips they would buy it; NOT using fab space is causing them huge losses on the foundry side. Yes, they could get more by selling to HP/Dell but they've already met that demand and can't sell more to them.

Also console CPUs are not as powerful as you think. The X360's CPU is about equal to a K8
dual-core [because I doubt it's more efficient per transistor than K8 and the transistor count is very similar]. A current $76 tri-core would far outperform it now; in 2012 a 32nm Llano quad-core with 480 shaders would be very attractive and be about the same price for CPU and GPU.

If anything, your argument applies to IBM too since they normally sell POWER processors for several thousand dollars in servers.

If Microsoft implemented AMD-specific extensions in Windows it would be anticompetitive and illegal. They wouldn't anyway, they much prefer Intel.

Don't even say Fermi, lol. You'll see when it comes out that it is hot (225W TDP, 190W typical power), underperforming (less than 20% faster than the 5870) and late (March or later). Its theoretical FP performance is a lot less than AMD's.

--

I've heard of PowerVR, but the performance isn't good enough. By July 2010 AMD will be sampling chips with on-die graphics equivalent to 480 shaders on a current chip; that's more than a 4670. PowerVR would be an option where power consumption and cost are more important than performance: Wii and handhelds favour this. While MS and Sony have yet to say keeping down prices or power draw is important in a home console.

Its not AMD who are below fab capacity its Global Foundries, just a quick correction. However the concept that they have fab space to give away may very well be coming to an abrupt end once they bring their graphics on board in both Fusion which will be a larger higher margin core with lower yields and their standalone GPUs as well as the global partners who want to use their high quality fabrication processes as TSMC has dropped the ball as of late.

The performance requirements of all console -> PC ports have been increasing with time as developers make better utilization of the Xbox 360 CPU. People used to say that you needed a single core to match it, then it was dual core and now a lot of games are starting to recomend triple/quad cores for games which were ports from the console. The ports aren't getting worse, the demands from the games are increasing. The Xbox 360 CPU has been specialised somewhat for games and out of order execution is of less benefit when code is directly compiled for a CPU itself.

IBM themselves don't have a huge desktop market to fall back on so they take royalties for development where they can. This is why they are more ameniable to console manufacturers because they need them from a software development and research and development standpoint. X86 doesn't need any more developers than it currently has. They also don't need to compete with themselves by giving away cheap mini computer components to compete with their low end desktop lines.

Microsoft on AMD extensions yep I see where you're coming from

Please don't write off Fermi before its come out. Its a pretty big leap to suggest it would only perform ~20% faster than Cypress when it hasn't been released yet. Theoretical FP performance is just that, theoretical. Theoretically a non unified PS3 GPU has more flops than a unified Xbox 360 GPU but in reality the latter outperforms the former. Whilst Cypress is a great GPU its an architecture which has been scaled from 320 stream processors to 1600 stream processors, from 600M transistors to 2,000M. So if the Fermi architecture is a good one it should be able to easily beat Cypress on effective flops and it definately has an advantage in flops vs bandwidth due to a wider bus.

i just hope they make the damn thing affordable, and get their first party games out within the first two years, gt5,god of war3 im looking at you.

The PS3 is currently still in last place and, last I checked, is still losing them money with each unit sold. PSP Go was a flop and PSP sales are dwarfed by those of the DS. They're billions of dollars in debt. Should they really be thinking about a PS4 right now?

There is absolutely no question that Sony has been working on concepts about the next console generations. They have the engineers and researchers in their labs and have certainly kept them busy. Work on any of the existing and future gadgets is a continuous operation - whether you think this and that is a flop or not.

More to the point: I think a logical step for a future console is a cell chip with 2-3 cores and an additional 2-4 SPUs (for motion stuff and other ideas). In 2-3 years, the 32nm fabs will be able to keep manufacturing costs at the level they are now for current cell chips. As for the GPU solution, a fermi solution is completely out of question due to the high transistor count and associated thermal problems (there might also be "philosophical" reasons for not going such a route). In designing a console, there are several ceilings (chip costs, max. thermal output) which exclude even trying to go certain paths, and attaining the level of high-end PC graphics solutions is simply not possible for consoles. The GPU situation is certainly the more difficult problem for Sony.