UE3 runs like shit on Xbox360 as well.

As a matter of fact, i have yet to play a PS3 game with such massive performance and visual issues as Mass Effect or The Last Remnant on Xbox360. Even the games made by Epic (the creators of the engine) have the same tehnical issues in their games, though not on the same level as the rest and this mostly because of the linear nature of their games.

The best thing a console dev can do is not use UE3 and that's a proven fact.

Using UE3 as a development engine still greatly truncates development time. Unfortunately that does mean it can be used for low quality shovelware (Hail to the Chimp used UE3).

But in terms of being an ideal solution for console game development, I'd have to say no. It just doesn't run as well as it does on a PC; even a modest PC with an up to date VGA card.

I think the reduction in development time may well be the best thing UE3 has to offer console developers (which is often enough).

Even so, Bioshock didn't have any significant issues on the 360; it just didn't native render any higher than 1280x720. Frame rates were never an issue either.

Like it or not thats the future and it doesn't stop there. Consoles are losing their ability to make games better on worse hardware. No more games that are made only in asm & stuff. In next gen there will be even more things that are handled by third party libraries and it will slow down whole game nicely. Why this doesn't matter on PC? Because thats what PC devs have been done since first 3D games. Also with different kind of architectures like CELL this will be more problematic, because you can't fix stuff that is made by someone else and if they haven't had a clue how to make it efficient on CELL its really bad for you. :)

@Greenmedic: I'm betting on Sony getting royalties. After all, Cell includes patents from all three. How do they split the money and what kind of deal they had considering the manufacturing, is a whole different matter. There's still a lot money revolving around processors.

@NJ5: It looks like Cell was one trojan in PS3. It introduces programmers into the code, quick drop in costs if it would have sold as expected (and the Cell developement would have ran as it is now), keeping production line running, while the costs would have been offset by the number of games sold.
Just think about the synergic advantages for a moment. Some people offer HDTV sales for the advantages, but that's the least of all that the PS3 had "brought home".

its been two years now since ps3 was released and it still lags with multiplatform games. cod waw ailiasing issues fallout 3 jaggie textures this proof also that the rsx is also hard to develop for.

I can't entirely agree since that's essentially saying there is no more headroom for consoles when it's a known fact that the most impressive games of any console platform invariably show up late in the life cycle.

The hardware doesn't change; developers simply become more familiar with the environment, coding and resource libraries are vastly improved, etc. Granted, there's no magical "unlock additional processing power" button in any console, but the best developers do tend to come up with the most ingenious solutions the longer they've been working within a closed platform.

It just means console developers can't take a PC hardware approach to development under the assumption that hardware improvements will eventually allow more people to run the software without scaling all the hardware intensive features back (Crysis being a prime example).

What will probably end up happening with the PS3 is that some (a minority) of its developers will be tier 1 developers that can actually utilize the platform to its fullest extent, while many will continue to make no special efforts to optimize and others will simply throw in the towel and abandon the PS3 as a platform (not as bad as it sounds, it just means less crapware from untalented developers).

Oh, i forgot:
@Greenmedic: Anyway it doesn't matter whether Sony sold/hadn't sold its plant, it still was planning to sell the CPU (or atleast profit from it). The reason they got rid of the facility, was because of its bad financials. Since i don't know the details of their deal, can't really say how the money is shared and who had right to manufacture and sell it and to who and how much.

It's one of those undefined issues that generally isn't public domain information. Plus there are the specifics of the individual patents each firm holds. Does Sony's specifically refer to application, or is it in the architecture of the chip itself? I'm guessing it's the former.

Either way, I'd have a hard time seeing Sony pulling in any significant source of revenue through the use of CBE chips in anything other than its own PS3.

Do they get a cut from every CBE used in a Toshiba HDTV? From each chip used in every IBM manufactured supercomputing cluster? It seems unlikely, but again, unless someone knows the specifics of the patents each firm holds, it's merely speculation.

@Greenmedic: If it includes Sony owned patents, it's very likely they get a share. Another thing is, that whether Sony have sold their patents. It also could have been, that Sony would have made money out of selling the chip from its plant etc.
I'm really having trouble believing that Sony would have just been "along", spending hundreds of millions(?) into the Cell developement and then let it "out to public" for others just to take advantage of it and for Sony to pay full price for the processor. Sony could have used, for example, four MIPS processors in PS3 and they still could be cheaper.

But each firm that holds a stake in the patents for the CBE (Toshiba, Sony, IBM) manufactures its own products that utilize the chip. The only reason Sony was initially in the processor manufacturing business was to meet initial yield demands for PS3 production. I'm pretty sure it was never part of their strategy to become a chip vendor for Cell processors.

As of now, they have no hand in manufacturing the chips themselves. With current production output from other chip manufacturers with the facilities to produce Cell chips, it is actually more cost effective to buy the chips rather than produce them in house.

To my knowledge, there aren't any other companies currently buying these chips than Mercury Computer Systems (defense apps, medical imaging equipment, servers) and their partnership is with IBM. Companies like Leadtek have plans to use it for video and audio encoding cards, but that's in the future. No specifics on who's supplying the processors.

But from the information available, the only company that seems to be making deals to use Cell processors in any equipment not being manufactured in house, is IBM.