I'm having issue with ESXi random crashing recently.

My setup (Not in WQHL, I know, but it's been running fine for 2 years up until now):

- AMD Ryzen Threadripper 3970x

- Asrock Rack TRX40

- Gskill Ripjaws V DDR4 32GBx4

- Samsung 970 EVO Plus 2TB

- Intel SSD

- WD Red / Gold 12TB

ESXi crashing sometimes as soon as VM started, up to 12 hours uptime. It's just random.

First occurrence happened right after my UPS battery restarted, which made the server went down.

When I booted it again, BSOD started to appear, with various messages when it happened:

- CPU Tried to Re-Acquire Lock

- Verify bora/vmkernel/sched/cpusched.c /

- #PF Exception 14 in world xxxx:vmm5:<random_VM_name> IP xxxx

- Spin count exceeded - possible deadlock

- Error closing the volume: . Eviction fails: Failure

- NMP: nmp_ResetDeviceLogThrottling:3839: Error status H:0x0 xxxx

What I did:

- Switched ESXi to different SSD and slot (by reinstalling)

- Cleaned up the case, re-applied CPU thermal paste

- Switched RAM slot and run only 1 single stick 

- Re-installed ESXi using version 7.02 - 8.01

- Unplugged every PCIe device, including Intel NIC and Avago HBA

- Replaced CMOS battery and adjusted date/time

- Reinstalled every VM OS

- Switched on only one VM at a time

- Switched LAN cable to other port

Attaching various message when it happened. 

Please can you help me. I'm running out of ideas.