With a virtual machine you can set reservations, limits and set the shares value.
Limits - you could limit the CPU that was available to a VM. A vCPU can only operate as fast as a single physical CPU core. With a limit you can prevent the VM from using more CPU resources than you want. So in your case you could set a limit of 1.5 GHz and the VM never run faster than that even those the physical cores could.
Reservations - with a reservation you can gauranntee a VM specific resources. So if you grant a 4 vCPU VM a 3 GHz reservation, 12 GHz of CPU capacity is used only for that VM. Downside is that if the VM is total idle, those CPU resources can't be used for other VMs even if they are maxing out the remaining CPU resources.
You can also create Resource Pools (RP) with shares, limits and reservations as well to better manage your host. For example you could have a production RP with your production VMs and another RP for test VMs. You might set a reservation on the prod RP or limits on the test RP.
You can also allocate shares. Lets say you have 3 VMs with the default shares (1000). If a resource is under contention, then ESXi will allocate resources according to the division of shares. So with 3 VMs with equal share values, CPU resources would be split 3 ways. If you were to allocate 1 VM 2000 shares, then under resource contention the VM would get 50% of the resources and the other 2 would get the remaining 50%. Shares only apply when a resource is under contention. If that's not the case then an individual VM could use all the resources it could get.
In your case it looks like you're looking to use Reservations, but as mentioned the down side is that idle resources that are reserved cannot be used for other VMs.