Originally posted by: Lonyo
But if they use the same system as some current solutions, where the IGP is used for graphics output all the time and the discrete just for 3D which is then sent through the IGP (I think that's what they do with some systems to save on discrete power draw when in idle mode), then it wouldn't make any difference.
But if they ran the display of discrete all the time and had an IGP CPU then it would be interesting to see if the IGP can be completely shut off, but is there any reason why Intel couldn't engineer it so it was? Current IGP's are integrated into the NB which is where the off-CPU mem controllers are, so why can't they do it when it's all in one package with the CPU?
Lonyo, I think you might be confusing it with the switchable graphics that's in recent laptops. With switchable graphics, the user can switch to whatever the graphics they want at the time(discrete for 3D, integrated for non 3D or battery life). Of course, it means if they don't switch it, they aren't taking advantage of the feature.
But normally when a discrete GPU is installed in a system with integrated graphics, the IGP don't function at all. Everything
including the display goes to the discrete graphics.
By Tokie
No, that's not what I'm talking about. What I'm saying is this: You have a 35W CPU, and a 5W chipset without IGP. You also have a 20W discrete GPU. Now, with Arrandale you would have a 35W CPU + 5W chipset (on/off die, whatever same functions) + a few watts going to the IGP because the IGP die is shared with the IMC. Then you still have the 20W discrete GPU. Now you end up with a higher TDP of a few watts?
That's actually a legitimate question, but I don't think it'll be that big of a problem. The vastly improved power management settings will work better in real world than simple theory will suggest.
Anyway, what's to suggest they won't use the power saving features to the full extent with the GPU/MC portion too?