Latency from core to core is pretty much always lower than processor to processor, because information doesn't have to be sent between additional devices.

For instance, sending data between four cores on one processor is easier and more efficient than trying to send the same amount between four processors with only one core each.

Windows doesn't treat multiple cores as multiple CPUs. Even in the Task Manager, it divides each processor into its own sections with monitoring for the load on each core.

Just a little mockup of what a single and double quad core setup looks like:  

Attached Thumbnails