High Availability is the restart of a virtual machine on a different host within a HA cluster, so doesn't need to be identical CPU as far as i know.
Design best practice would suggest each node of the cluster would mean each host is near identical, same cpu version and family (but maybe different speed / cores etc).
Check out the vSphere High Availability (HA) Technical Deepdive - Yellow Bricks by Duncan Epping
HA will still work even with completely different CPUs between your physical hosts. You could even mix AMD and Intel CPUs, though I certainly would not recommend this.
This is because HA does not rely on vMotion (DRS does) to perform live migrations, HA is just a re-power on operation from a hard powered-off state (think plugging the power cord).
This begs the question of why you don't want to use live migration on your clusters. Remember, HA has the same shared-storage requirement as well.
Agree with the others, it should work regardless.
HA will work. if using different CPU families in cluster then enable EVC mode for compatibility.