When I've had them on my PC it was always heat related. I have a 5800X3D on a 360 AIO and Strix 3080 rig. In an 011 Dynamic case with plenty of cooling. I also have a bunch of LED prettiness on it controlled by iCue. VR hits your PC a lot harder than just running on a monitor will. The only time I had a lockup in AH, which generally doesn't tax it too much, was when there was an update to iCue that stopped using the hot spot sensor on my GPU to ramp up all the fans. Fans didn't speed up, box got hot, it locked up. That got fixed in another update and problem solved.
Even so, I've had it happen in other games that do put more of a hurtin on my rig like IL-2 and MSFS. Running both the CPU and GPU at stock, they draw a lot of power and so generate a lot of heat. I found that by slightly under volting both I'm getting the same performance wth less power and heat. This CPU can't be effectively over clocked anyway and will tend to throttle when it gets hot so it benefits from the under volt by keeping it in the boost clocks all the time. Took longer to find the voltage curve on the GPU. Taking away too much power is like running an engine too lean. It'll just lose it the other way. Tiny increments and testing. After the tweaking I haven't had an issue.
But ya really have to watch your temps voltages and power to find the right balance. I use HWInfo and Afterburner for monitoring, 3DMark suite and Cinebench for testing.
Sent from my SM-G991U using Tapatalk