The incentive to buy next generation consoles will have to be something other than raw processing power ...

The "general leaps" in graphical processing power have also included "generational leaps" in the cost to develop games. Since the NES each generation of game has resulted in development costs increasing to 2 to 4 times the level of previous generations. Being that games already need to be released on the XBox 360, PS3 and PC to get multi-millions in sales to be financially viable, games would (probably) need over 10 million sales in order to justify the increased development costs. While there are a handful of games a year that could justify this increase in cost based on sales, most studios are (probably) going to try to keep costs close to where they are currently; which would limit any increase in the quality of visuals.

While additional processing power could be used for "techical" features, like 1080p output (or 3D display), producing a console at this point in time that trys to sell itself on processing power alone is a losing strategy.

Obviusly power alone is not going to do anything, it is a tool for developers to create new experiances that couldn't be done before, just like new controllers etc. My point is that next generation needs something like that to suceed, current consoles but 4x more powerful is not a recipe for sucess, as that is not going to mean experiances drastically different from what is available now. But it's not like MS can launch with Kinect 2.0 to make up for a less that staller leap in power as the kinect experiance is already available on 360. So unless someone can get a BMI working in time for next generation then a generational leap in power (~8X) is the obvius easy route. 

As for development costs they are increasing despite static hardware, average AAA games today cost more than AAA games 4 years ago despite being on the same hardware (there are outliers like GTA 4) and adjusted for inflation Shenmu is still one of the most expensive games ever made.  If developers want to lower development costs then they need to change how they make games. With a staff of ~90 people CD Projekt RED managed to make a AAA 30+ hour RPG with 2 completely different middle sections for 8 million dollers yes the average wage is ~50% lower in Poland but Final Fantasy XII cost $48 million on the PS2, FF IX cost $40 million on the PSX, Too Human cost over 60 million etc and the Witcher 2 has far better graphics than any of those games and cost a fraction to make, Crysis cost $22 million yet Killzone 2 cost $45 million and Halo 3 cost $60 million and COD costs ~$40-50 million a pop (with a $100+ million marketing budget). As you can see better graphics don't always = higher budget. Hell your average marketing budget for a AAA game is usally $20-60 million usally higher than the development budget. The biggest factors are scope and complexity with nevelty also a big factor (making something that has never been done before is expensive) that is not to say that graphical fidality is not a major factor but it's not the be all and end all for the increases in development budgets it's more a facor of fidality x scope + novelty. A strong push towards higher margine digital distrabution will also help to offset production increases next generation. 

We will also likely see far less custom engines next gen as they get ever more complex to make and engines like Unreal and Cryengine get ever better. Also a move to fully dynamic lighting and shadowing which will mean less expensive precomputing of effects, hopefully more procedural generation (why stop at trees Speed Tree? where is Speed Terrain, Speed NPC etc?), more of a focus on image quality over graphical fidality (1080p native with AA and never droping below 30fps), tessilated LOD systems meaning that less LOD models have to be created plus less popin and also more stylized graphics there are plenty of art styles that look good yet require simpler models and textures, not all games need to look realistic, but that shouldn't mean everything is cell shaded ether diversity is key. 

It's unlikely that there will be a substantial difference in power between the two next gen, just as there wasn't this gen. It's also unlikely that it'll really matter, just as it hasn't this gen. And given that both companies suffered significant losses as a result of their hardware strategies this gen, particularly Sony, which probably won't ever be in the black on PS3. 

But perhaps more importantly, there doesn't *need* to be a tremendous leap in power for the coming generation. Very few people are complaining about the visual quality of games such as Uncharted 3, Gears of War 3, etc. We've finally reached the point of "good enough" graphics for most people. So where does that leave the next generation?

Simple: refinement. They should evolve the existing tech; Cell and Xenon with more cores, updated architecture to hopefully support features for improved AI. More recent graphics chips; DirectX 11 compliant for Xbox, current OpenGL nVidia for PS. More RAM--say, 2-4GB. Increase system bus to 256 bit from 128 bit. Refine the controllers (though 15 years in, I doubt Sony will change much with the DS design).

By doing this, they'll retain backward compatibility with current games, keep hardware design and construction costs low (we don't need another $5-600 launch debacle), allow developers to keep leveraging the tools and techniques they've developed and refined over the last 6 years, and keep compatibility with all of their existing software and services. 

Meanwhile, they can reasonably drop the existing systems to $149 or less and support them during the transition period, just as Sony already did with PS2 while PS3 struggled to gain a foothold.

At the end of the day, though, there's an interesting truth at work here: MS and Sony have been forced by Nintendo to at least consider some action. There's no guarantee we'll see PS4 and Xbox 420 announced this year--but there's a pretty decent chance.

You're definitely the one who's correct. In spite of what the other poster is saying, PS Vita IS using "off the shelf" parts. It's running a quad core Cortex A9 ARM processor (we're seeing them appear in smartphones already, starting this week at E3, and the dual core variant already exists in the iPhone 4S, iPad 2 and a fair few Android devices). It's also running a PowerVR SGX 543MP4+, which is merely the quad core version of the dual core PowerVR SGX543MP2 used in the iPad 2 and iPhone 4S.

I've been a systems engineer since 1996; I'm fairly hardware savvy as a consequence of that experience and training, and what struck me immediately upon the PS Vita's announcement was how generic the core hardware was. I don't say that as an insult; I think it's the smartest hardware design move that Sony has made in many years--one that I hope they replicate by developing PS4 as simply a more refined version of PS3, with an updated GPU, more cores on the Cell CPU, more RAM and a wider system bus.

PS3, like 360, is a very capable machine to this day. There's very little to be gained by an expensive complete re-imagination of the system, and a lot to be lost--including the library of existing AAA content and a slew of services, not to mention all the content now available on both PSN and Xbox Live. Microsoft and Sony would both be very, very foolish to release systems so radically different that they lose compatibility with that massive library of downloadable content :).

I remember reading that the SGX543MP4+ had been customised by Sony. But I guess I could be wrong I can't seem to find the article now. But yes it is definatly the least exotic peice of hardware Sony have ever made. It will be interesting if Sony decide to use that stratergy with the PS4 or stick with the same tech they have like you say and use an upgraded Cell and Nvidia GPU. If the rumours of "Loop" are true MS seem to be sticking with a similar stratergy as the 360 with a multicore IBM CPU (I would be interested to know a systems engineer's opnion of whether they are likely to stick with a PPE bassed in order CPU or upgrade to a Power 7 bassed out of order chip) and a powerful AMD GPU (presumably with a lot of EDRAM) but then they screwed it up by seemingly moving to a split RAM pool to save money.

I don't know a great deal about Power7, but I've read interviews with IBM engineers who've said that they've taken lessons learned from Cell and incorporated them into the Power7. We know already that Wii U's processor is based on the Power7, and yet still retains backward compatibility with Wii games, so it's probable that they could do the same with PS4.

Given the critical importance of the SPU's in Cell, which developers have to code specifically for, I'd think that any new variant that incorporates Power7 core changes would still need to keep the SPU's. It's possible that they may be able to adapt the out of order execution units to the primary core and keep the SPU's (maybe even adding a few more), or they could introduce logic to map calls for SPU functions to equivalent (or alternatives) on a more native Power7 chip, but I'm not sure what would be the more cost effective (or compatibility-effective) method.

Looking at the huge quantity of content now available on both Xbox Live and PSN, though, I just can't see them moving to completely new architectures right now. There's just too much work that they'd undue if they lost compatibility with the past 6 years worth of software that's already deployed. Before PS and Xbox are anything else, after all, they're *products*, and businesses sell products to turn a profit. They know there's money in this content library, but only if it continues to be usable once today's consoles are yesterday's news.

I hope you won't see that as a bad thing, though, I certainly don't intend it as negative judgment :).

I didn't read this yet but I am familiar with PSM. In my closet, as we speak, I have an issue of PSM. The EXACT SAME MONTH that the PS2 was revealed in the Official Playstation Magazine--with pics and all, PSM reported from a reliable source that the PS2 would:

-Have four controller ports
-Look similar to the Dreamcast
-Allow you to receive messages from friends about games
-Have an inbox that gave you "missions"

I can dig it out of my closet for specifics but PSM (which was headed at the time by current Nintendo Power head hocho, Chris Slate) isn't what I would consider the most reliable of sources. And on top of that, what's the big deal with being the most powerful? Has that ever mattered in the world of video games? Has the most powerful EVER won (Answer: No. Even the Snes beat the Neo Geo)?

And even if we get these specs now, remember how:
-Kinect has its own processor but the final design was cut
-The N64 was supposed to produce CD quality audio but in the end, we got only snippets of music in comparison to PS1
-The Wii was supposed to be so much more powerful than an Xbox
-The Cell was supposed to be more powerful, with the more consoles active (I still don't know what the fuck that meant)

Take this article with a grain of salt. I've seen this too many times before.

With games turning more and more into a service industry and with the massive sunk cost people have with PSN and XBOX Live I imagine that consumers would definatly prefer the ability to bring all their digital content over seamlessly and it's not like software emulation for the Cell is going to be arround for a very very long time so it doesn't seem like Sony have much of a choice in the matter when you think about it. 

Hmm combining Power 7 cores into a cell design does sound like a viable option to get the 4 SMT threads per core and OOO execution and keep backwards compatability, I am sure most developers would prefer more traditional cores over lots of SPUs especially if the console uses a modern GPGPU. Or I guess they could just resurrect the old cancelled 4 PPE 32 SPU design that IBM was planning. I suppose if they wanted to maintain hardware level backwards compatability they would also have to keep using expensive XDR RAM as DDR3 still hasn't matched the bandwidth XDR offers (on a single channel anyway and I don't think they are going to use tri channel RAM on a console), and I don't think anyone has actually even licensed XDR 2.

Yeah.

We're hitting a glass ceiling in terms of development costs, graphical quality, power consumption, and first party  budgets . Next-gen design will be about efficient and well-priced  consoles.

Now this I can agree with.

Hopefully they make them streamlined as well, easier to develop for, better user interfaces and less DLC and patches. I want to go back to basics.