Machiavellian said:
Yeah, you are right that duel GPUs are nothing new but how many duel GPUs in an APU are out in the market. How many combination of an APU with a 2nd GPU do you see out there now. How many of such parts does AMD make now. I really do not know what to say. You are really not thinking your argument through very well. How is MS going to make a statement about a design that does not exist today an not tread on AMD announcement. You act as if things just end with that one statement and nobody is going to ask How, where is the GPU, what are the specs, how does it integrate with the APU. So MS release that Pandora box and you believe they can just say No comment from that point on. You do know that the hot chips conference isnt an MS conference right. You do know MS cannot delay the conference because they want to announce something later in the year. You do know that MS presented at Hotchips for the 360 as well so they probably had an agreement with Stanford to present the X1 chip. How is showing the motherboard would break the NDA when the dGPU is rumored to be on the second layer? If you are going to argue a point, at least understand the technology first. Better GPUs at what time. As I mentioned without knowing when the actual APU custom design was built and finish who knows what was avaliable. Have you ever setup a PC using crossfire or even Nvidia SLI. depending on the card, you can actually get better performance with 2 GPU cards than one uber card for cheaper. Case in point would be the Nvidia Titan. You can put together 2 680s and for cheaper than one Titan and in most scenarios get the same performance or better. This is not a car, there are no moving parts. That analogy doesn't really fit with the hardware in question. If that was the case then having a CPU 8 of them and one GPU would make the whole design really unstable than just adding another GPU. The question is not why going from a mid level GPU to a high end, the question is what does the gGPU does that AMD decided to create such a part. The way your argument is structure it appears as if you believe MS built this part but its AMD who built the part. Why does AMD feel the need for a dGPU, what is the benefit and cost savings compared to a regular APU with a GPU card. What would a customer purchase a APU with a dGPU over other combinations. The question isn't why MS wouldn't use such a chip because its obvious, the question is why AMD made such a part and what are the benefits. We will not know that until the 29th. |
It really depends, did AMD build a stacked APU themselves? I personally don't think they have the engineers capable of that. But assuming they did, why hide it? Why big up the PS4 over the Xbox One (Refering to the Huma tech debacle) if the Xbox One is as powerful as a high end video card? Which means, MS made the changes. If MS made the changes why not tout it like the coming of christ? Why bother saying you've increased GPU and CPU clocks when you've never actually given any clock information? You could easily have said "Hey this is the GPU clock, it's higher than the PS4s" vs "Hey we increased the clock from a base clock we've never announced". Why hide a second GPU if you've developed the tech? AMD would be under NDA not MS. Unless they literally added the GPU last second, which ... well I wouldn't wanna be an Xbox employee come November when Xboxes start melting.
On the flip end, if MS is under NDA, I'm sure AMD would be pimping out their tech and putting it in their roadmaps, which they haven't. For a comparison; they've mentioned they're going to be manufacturing PS4 APU's for consumers.
Which leads to the next possibility; dev kits. It's quite possible that the devkit has a second GPU for development of assets, or debugging. It would make those crazy mistranslated e-mails make sense. It would also make sense that Major Nelson would be confused cause he was looking at the devkits and thinking the final build would have a second GPU.
