Well, at least it sounds like Gears 2 will be awesome....

Geldorn said: Squilliam said:  @SPE code usage - " On a Pentium 4 HT running at 3.4 GHz, this algorithm is able to check 24-million edges per second. On the Cell, at the end of our optimization, we achieved a performance of 538-million edges per second. This is an impressive result, but came at the price of an explosion in code complexity. While the algorithm in Listing One fits in 60 lines of source code, our final algorithm on the Cell measures 1200 lines of code."     That's a bit of a false comparisom though. Not all algorithms will need this sort of optimisation. Nor is it said that you can't just run the 60 line version for similar (i.e. about 24 million edges per second instead of the Cell powered 538 million) performance instead.

Thats true, I have seen but cannot show that the Cell CPU requires a larger overhead (not massive) than the Xenon because of how the cores are arranged and the number of cores. Essentially thats what I was getting at.

 

Threads full of numbers that mean very little are great.

Sod opinions. Who cares about opinion when numbers are so much fun.

@ c0rd

Although I agree with regard to the endless threads created by 360 fans about the minor early difference with regard to 360 to PS3 conversions. IMO the long term differences are pretty big, Blu-Ray, default harddrive as well as the Cell processor make quite a difference with regard to long term potential.

Of course with the 360's headstart and the PS3 being very differently architectured than legacy systems in some crucial areas (requiring game engine redesign and imposes a small learning curve) the already demonstrated differences in current games are smaller than will be the case for the long run. (Hurdles I already talked about years before the PS3 hit the market)

And I agree with you about any kind of game can be done on any current generation console, but with sacrifices, not only graphics (anti-aliasing, resolution, polygons, texture quality, etc) but also with regard to things possible to do on screen at once and at what performance, complex AI, world size, world complexity, physics, sound quality/complexity, etc, etc.

Of course like Squilliam pointed out, the more resources available and more features implemented, this usually leads to far more complexity. For instance the Amiga provided custom chips which were extremely powerful for their time and provided pre-emptive multi-tasking, although it was often still possible to take simplified approaches within a couple of years game engines and programs became a lot more complex compared to similar less impressive versions written for single tasking systems lacking such advanced custom chips (competitively push games and programs ahead of rival offerings).