Hi Scepter,
As people have mentioned this "can" be caused by faulty hardware but it can also be faulty software. The ESXi build version you are running is GA, i.e. General Availability and therefore is quite old. I would suggest that you schedule some downtime (assuming you have some uptime :smileywink:) and apply the latest patches. There are a number of VMware KB articles related to this message and one of the newer patches may resolve this issue as based on the PSOD it appears ESXi was migrating the vCPU to another pCPU when the exception happened or was experienced.
Anywho make sure you have a good recovery path especially as you appear to have customers relying on the hosting layer. I'm hoping that you can at least vMotion the workloads to another server as you patch this one?
Also make sure that the firmware/BIOS is up to date on the hardware as well. There are a number of instances where this can be resolved with BIOS/Firmware upgrades as well. Same advise as above - make sure you have good reliable backups or a known recovery point.
Kind regards.