No kidding, Microsoft simply wouldn't be able to control costs next gen if they went that way again. Xbox 1 was a total loss, they never earned a single dime from their debut console and they started the following generation more than $7 Billion in the hole. They will end THIS generation (as of the last fiscal Q) with more than $8 Billion in the division's bank. That is a complete turn around, I can't imagine them abandoning what worked so successfully for them just to get back in bed with a company that they needed arbitration in order to settle their differences.

I'm really stunned that Fudzilla had this whole thing nailed down as early as April, they stood by their sources and everything they said has lined up. They were the first ones to break the six cores and 8+GBs of memory. So if THAT turned out to be true, then does this really mean that their claims about two GPUs are also true?

When I first heard the two GPUs claim I laughed it off, but it looks like a dual configuration that's trying to match the higher performance of cards that will be coming later might be what this is really all about. Oddly enough it actually makes alot more sense for Microsoft to go this route then waiting until 2014. And for what?! Stuffing a brand new and consequently BURNING HOT game card in a small shelf component console? We all knew that wasn't going to happen. Oh Microsoft you clever dog, you're going to do the same thing you did with the Xenos/X1800-X1900 aren't you?

Hmmmm......I'm actually starting to become a believer here.

I can't wait for Xbox M8.

The two gpu thing is likely to just be their best attempt at simulating what they expect the final custom, single gpu to be like. Then again, perhaps another way to look at this is that their the goal is to use one GPU to handle the majority of the load as has always been the case, and then have the other GPU, perhaps just as capable or -- more realistically -- a GPU that's somewhere in between slightly less capable and much less capable, handle specific graphics oriented tasks, designed to relieve the main GPU of certain tasks, potentially allowing developers to cheat and aim higher than they might otherwise be capable of doing if the main GPU was by itself.

I guess think of it as being similar to how devs tend to utilize the CPU in ways to try and take some of the load off of the GPU, or how the Cell's SPE were used to take some of the load off of the RSX, except in this case, there will be a weaker GPU to accompany the main GPU that can do pretty much the same thing, only better, and then devs can still in addition to this continue to take advantage of using the CPU to help with certain tasks. For what you'd lose by not having one monster GPU, you could probably more than make up for by intelligently lessening the workload on not one, but two GPUs, while continuing to get assistance from the CPU.

Although, I'm not sure if I'm expecting a two gpu setup, as there are apparently development drawbacks to properly utilizing both gpus efficiently enough on different graphical tasks or whatever. I guess it would depend on how good the development tools end up being, which Microsoft seems to have a pretty good track record with.