Haven't had the time to perform enough testing yet to be 100% certain but hopefully I've identified the real culprit and addressed it.

As I was thinking from start, this kind of issue where the computer physically and out of the blue looses power and immediately is switched off...that kind of issue can't be caused by software. If it was, that would be one very nasty piece of code!

So, I started thinking and collection information and found other people having the same kind of issue when playing other types of games. Was good to see it didn't appear to be a X-Plane issue even when I only experienced it when running X-Plane. I also learned the majority of people having this issue were using a 2080 Ti just like me. The 2080 Ti is quite a power-hungry GPU.

I then came to think of my power supply unit which is a Corsair HX1000i. Certainly powerful enough for both my 2080 Ti as well as my 8086K overclocked to 5.1 GHz. But...with the PSU by default running in Multi Rail OCP mode meaning it will limit the amount of power that can be delivered on any single connector/cable...now that can turn out to be an issue when using a standard PCIe cable to feed a 2080 Ti.

This since the standard PCIe cable normally is a split cable meaning the same cable (and thus connector on the PSU) will be used to feed both power connectors on the 2080 Ti. As soon as the GPU asks for power exceeding the limit set by the Multi Rail OCP mode, the PSU will immediately switch your PC off in order to protect your components. So, what I initially thought was a problem most likely turned out to rather be a safety feature built into many modern PSUs mine included.

There are two ways to sort this situation.

The quick and dirty option is to change your PSU from Multi Rail OCP to Single Rail OCP. This allows the PSU to deliver the max amount of power it's capable of on any of the connectors/cables. This would fix this kind of issue since that would mean regardless how much power the GPU asks over that single PCIe cable, the PSU will deliver. The bad thing is you might end up with a cable or connector that could overheat since they aren't designed for that amount of power delivery.

The second option which should be the preferred one for anyone for safety reasons is to keep the PSU in Multi Rail OCP mode and instead plug in an extra PCIe cable meaning there will be one separate PCIe cable per connector on the GPU leading to it's own separate PCIe connector on the PSU.

This way, the power delivery on each contact and cable, both on the GPU as well as the PSU, will be split 50%-50%. This means the risk you would trigger the Multi Rail OCP is close to none as long as you're using normal components in your rig and not feeding them with insane amounts of power such as when doing extreme overclocking going after world records 🙂

I think the reason why we haven't heard that much of this issue yet is twofold. Not that many people have 2080 Ti GPUs yet and of those who already got one, a certain amount is using a PSU powerful enough to feed the 2080 Ti but lacking Multi Rail functionality. Meaning their PSU is constantly running in Single Rail mode willing to deliver all its power on any single power connector/cable.

For that reason, this group of people won't notice the issue I've been having. What might happen though going forward is the amount of people reporting cables and connectors physically damaged by heat will increase. Without them even understanding why until it's too late since everything seemed to work just fine...until their PC started to smell burned plastic or worse...

What really pushed me in the right direction was watching this video where a very knowledgeable guy from Corsair explains all there is to know about Single Rail vs Multi Rail. He does say (IIRC) the risk having issues running in Single Ray might not be that big after all.

However, if you want be to on the safe side and you're careful about your components, you certainly want to use a PSU capable of Multi Rail OCP and make sure that is the active mode. And considering what you pay for a 2080 Ti for example, I think anyone not willing to do whatever they can to protect that piece of hardware would be plain stupid!

Here's the clip if you're interested.

Sorry for the wall of text but felt I wanted to give you a picture as detailed as possible since I'm guessing this information could be useful for you if/when other people are reporting similar issues.

So, I'm quite confident to say you can rest assured there's nothing wrong with X-Plane. At least not in any other way than X-Plane putting your hardware up for a real challenge enough to trigger the Multi Rail OCP to go off shutting your system down to protect your valuable components 🙂

Many thanks again for your very prompt response when I initially reported my problem as well as for following up now.