Does Cloud Computing Need a New Generation of Chip Technology?

My immediate reactions to this question are "Why?" and "No!"... Maybe other responses will make me think some more?

Well, cloud data centers are already huge, sprawling complexes (Amazon, Google, Microsoft, to name a few) that need a lot of power and real estate. Demands on these and other data centers will only keep scaling upward. They certainly will need faster chips with greater capacity, if nothing else. And the more this can be compacted, the better.

Cloud computing is based on the core tenet of infrastructure that is highly scalable and always available. The core architecture calls for a scale out distribution of load. Current chip technologies are more than suitable for these constructs. In fact the volume of units that would be required in a utility cloud asks for more commoditization of available chips. Some areas do not need innovation specifically related to hardware assistance for running hypervisors on the bare metal chip. This would make the virtual machine run faster and further scale up the cloud. Another area would be security provisions at the hardware level that work along with the hypervisor.

Sure - how about standard and more mature virtualization extensions for the x86 architecture? Then, enterprises don't have to pick AMD-V or Intel VT and exclude the other vendor in the process for the sake of portability. As well, we can stop needing software for effective virtualization. I expect this would ultimately improve security due to the elimination of the software.