First of all, 360 is at 60 not 100, and yea PS3 is at 30. But it really has nothing to do with the strength of the consoles, its just that EA has become a bunch of lazy bums because they have the only NFL game and they know it. PS3 isn’t harder to develop for, its just different, all they have to do is break down their code into 256kb bits so that the Cell can use it, seeing as how the SPE’s can’t see the local memory. And I can tell you, as I am a bit of a techie, that the strength of its archetecture is far superior to that of the 360, its Cell processesor is going to do some crazy things.[/quote]
And what does this mean in english?[/quote]
Basically that while the Cell Processor runs circles around 360’s Xenon processor in any number crunching application (I’m talking 20x faster minimum), it has a new memory model that developers have to learn how to use, and that can be a b**ch.
Traditional x86 processors utilize on core and a local cache. The cache is a small portion of extremely high speed memory that keeps the most common commands the processor uses in the memory, so when that processor needs the code for that command again, there is less latency time while it searches for it in the much slower system memory. Cache speeds things up alot.
Think of this as one really big lineman with one play, stopping an entire defensive line.
Then you have dual/tri/quad core processors, which are multiple cores floating on a single cache. These are slightly more difficult to program for, because it forces developers to break their code into several different parts that they feed to the different cores. However, because they operate on the same cache, it is not that much harder to do.
Think of this as 2/3/4 smaller lineman with one play, stopping a defensive line. (This is the variety that the 360 uses)
Then, you have the cell… heres where it gets tricky to explain to non techies. The cell is one standard x86 processor linked on a high speed bus to 8 SPU’s, or Synergenic Processing Units. These SPU’s are extremely fast in number crunching operations, but fail in some normal game functions which is why they never emerged mainstream. The one standard x86 unit, or PPE, has a cache, and can see the main memory. However, the SPU’s cannot see the memory, so they cannot on their own retrieve game code. What the PPE has to do is constantly stream information to the SPU’s, which then crunch the info in much smaller portions. (Which is why they are synergenic, they take an operation like (5+4)*(3+2)/2, and break it into smaller chuncks, where one processor SPU figures 5+4, another figures 3+2, another multiplies the sums, and another divides by 2) In a perfect enviroment, the PPE is streaming game code to the SPU’s quickly, who are processing it piece by piece, and spitting it back out to the PPE who can take that processed information and pass it to wherever in the system it needs to go next. This is easier said than done, however, it is a completely new enviroment for developing games, and some devs have trouble justifying all the time spent on learning the new format.
To finish the anology however, the cell is like having 7 small yet quick lineman stopping a defensive line, with one “Commander” lineman in the backfield constantly changing the play on the fly to adapt to the defensive line.
Hope someone out there got that lol…