Looking at the way mobile CPUs have been developed in the past, it's very clear that this isn't the right way to solve the problem of producing a high performance, low power CPU.
AMD or Intel would simply take a desktop CPU, scale it down (in both frequency and voltage), make it go to sleep as much as possible and call it a mobile processor. This worked for a while when we were dealing with relatively low power desktop CPUs like the Pentium III, but take a look at the power requirements of a mobile Pentium 4 running at 2.4GHz and you'll quickly come to the conclusion that there has to be a better way.
Now look at AMD's plans with their mobile Athlon 64; in this case they're taking a server CPU that was already scaled down to be a desktop part and doing whatever it takes to make it a mobile processor.
Neither AMD nor Intel's approaches to mobile CPU and platform architecture will work, and it's very clear the reasons why; other than reducing clock speed and attempting to keep voltages as low as possible, the CPUs are still architecturally identical to their desktop counterparts. There are always going to be parts of the chip that aren't optimized for low power operation, making the CPU far from well designed for mobile applications. When the architects at AMD and Intel are designing their desktop or server processors, they aren't concerned with battery life on a notebook, and thus the vast majority of the decisions they will make will be insensitive to the needs of a mobile CPU.
Moving beyond the CPU, the chipset and rest of the platform are also taken directly from the desktop world. The North Bridge in today's notebooks is identical to what we see on desktops, and with the amount of traffic that goes through that chip, there's definitely a lot of wasted power.
The amount of inefficiencies in current mobile architecture is astounding when you think about it, and after today, you'll understand how they're all going to go away…
