CPU frequency governor can scale the CPU frequency up or down in order to save resources.
In a physical server usually have sense, but I don't know if this is applicable for a virtual guest running under VMware.
Does make sense in a custom kernel avoid the option CPU Frequency scaling?
CPU Core power management is a function that can be owned by the system/BIOS on its own but usually it's a collaborative effort with the OS, in an Intel world this is done via Speedstep extensions, AMD have similar mechanisms. VMs each have a 'CPU mask' that the hypervisor exposes to each VM to tell it what CPU capabilities it can call on, some of these are core power management related - these all link into the 'EVC mode' settings of your clusters. Configured correctly this allows the OS inside a VM to tell the hypervisor how much resource is required and when. Now obviously the hypervisor doesn't just do to its resources what a VM tells it to but it does take it into account when scheduling its resources and it's perfectly reasonable to assume that if lots of VMs are doing little CPU work and they tell the vCPUs to slow or sleep that that might influence the hypervisor to do the same, but it is an influence - not a direct result. Is that clear?
