PC Reboot with MSFS Only

Bug Reporting Hub Crashes (CTDs)
,
1

Edit: Testing is still in progress, see the attached “solution” for the latest updates. The reboot issue appears to be solved as of now, but CTD w/ driver timeout is now happening.

Original Post:
I am genuinely at a complete loss here, so I thought I should finally turn to the collective mind.

I am experiencing hard crashes with MSFS (no, I wish it was just a CTD). The sim would just randomly stop, and my PC reboots (most of the time unsuccessfully until I force a shutdown, then power up).

I have run just about every hardware stability test there exists, and my PC passes AIDA64, Prime95, OCCT, Furmark, you name it. (Edit: MEMTEST with 145,000% coverage was done with 0 error.) I have tested every component independently and collectively. I have used P3DV5 and XP11 on 16-hour overnight flights with no problem. Yet, MSFS just randomly reboots my PC. This is with the MSFS running completely stock, with not even Marketplace addons like the WorkingTitle NXi installed.

My hardware is all practically brand new (less than 6 months old), and they are:
AMD Radeon 6900XT
AMD Ryzen 9 5950X
Corsair 4x8GB 3,600MHz CL16 RAM (tried running at 2,666MHz stock speed as well)
ASUS Dark Hero VIII
ASUS ROG 1000W PSU
None of them are overclocked. (I suspected overclocks causing a problem, so I disabled them all, but it didn’t help the issue.)

I’m genuinely at a loss for what this issue can possibly be. The two common suggestions I’ve seen so far are PSU failure and RAM problems. My PSU is almost definitely working properly since AIDA64 logs show no power demand spikes at the crashes, and my 12V, 5V, and 3.3V rails deliver 11.89V, 5V, and 3.31V at full load, respectively. (Not to mention it’s literally brand new from a reputable brand.) The RAM is most likely working correctly as well since they are run at the 2,666MHz stock settings following suspicious about DOCP not running correctly. I have tried plugging the PC into the wall directly, multiple sockets, new cord, etc.

What’s even more strange is that, sometimes, when paused with the Esc key, and alt-tabbed out, the sim would just crash in a few seconds. (Actually, make that a crash like that every time I pause with the Esc key.)

At this point, I’m not even frustrated, I’m genuinely just confused. So thank you in advance to anyone who can give some suggestions.

2

Dated article but still very relevant. I used it around SU3 and SU4 when I was getting one CTD out every six or seven flights. It might give you insights.

As for RAM, OC is only one possibility. What that doesn’t tell you is if one stick is faulty. For that, old fashioned pull it off the system and run on one, then swap.

2 Likes
3

System resets are usually caused by PSU issues. My issue was that the PSU couldn’t handle going from light load (menu and such) to getting slammed on the 12V rail when actually getting into a flight. My 2080Ti and 10940X would go from like 300W to near 850W total system power draw and the PSU over current protection would trip. I replaced my seasonic 1000W prime ultra titanium with a 1300W prime platinum after an RMA didn’t fix the issue and the problem went away immediately.

1 Like
4

Also

And heat issues…anywhere…but these can be hard to track down if its not something being monitored. Ssd can and do overheat, ram and m.2 can be very susceptible as they tend to be very close to motherboard and tucked away from airflow

2 Likes
5

I have a very similar sounding problem - although most of the time I CTD - but I do get some lockups that result in a reboot.

I’m running a new system- 5950x with 32gb RAM at 2666 on MSI B550 Carbon Gaming mobo. Windows 11. Nvidia 3080ti. 850W PSU. No OC. Nothing in the community folder.

I can’t fly longer than a couple of minutes without a CTD. I also get CTDs reliably when alt tabbing out of the sim, when ending a shadow play recording or often just going back to the main menu.

I’ve also spent days trying everything on the forums. The only things that seem to improve it slightly was power limiting the GPU (to approx 60%) and ensuring XMP was disabled for the RAM. This helped a bit but hasn’t fixed the problem. I also saw the same probs on Win 10.

Wonder what else our systems have in common!

6

I think you may need a bigger (or better) PSU although draw can often be mitigated just enough by unplugging unused USB equipment and unplugging optical drives etc. If that’s 4x8GB ram then removing two sticks shouldn’t harm performance if you use your MB’s preferred slots (2&4 on my MSI board) although I can’t say the same for 2X16GB as performance can be affected by running single channel.

7

Power spikes is a BIG issue with MSFS apparently !
I would have expected spikes in games with very high pace action - not some real-time sim flights.

8

Hello @sunshineyonder ,

if other recommendations fail, I would recommend to disconnect other power hungry items from your power outlet/usb ports.
For example, I had my PC and quite a few monitors etc. connected to the same power outlet. When I additionally connected a force feedback steering wheel to the same power outlet/power strip, the complete PC crashed/turned off instantly after some time playing a game. I do not know if I just pulled to much power over one outlet, or the inconsistent power draw caused some kind of safety mechanism to trigger in the PSU/Mobo. If this could apply to you and other recommendations fail, try to use multiple power outlets and try without peripherals like controllers to ensure it is not a power draw (potentially also from usb) issue.

9

I’m sure there’s reasons and further optimisations to come however that reminds me of a thunk I had a while back.

This is a flight sim and in many areas it’s coders have a practically unlimited supply of latency to work with as a quarter or even a half second delay for most aspects would go unnoticed by us simmers, I dare say this includes pilots of the twitchy Extra and Pitts too. It’s not always about speed.

1 Like
10

I have taken a look at this tool. It does show me the repeated LiveKernelEvent errors.
I also ran a Windows memory diagnostic test, and a 145,000% coverage MEMTEST overnight. Both returned no errors.

11

If I understand correctly, I may be running out of power on a 1000W PSU due to an instant demand?
This would make sense if it crashes immediately after loading a flight, but none of my crashes are anywhere near a demand surge. The last logged wattage for the CPU and GPU are both average values. In fact, in the places where the power does spike, I’ve never had a crash.
I will be looking at the USB-induced power shutdowns suggested elsewhere, but my peak power demand when in the sim has never exceeded ~700W, based on the AIDA64 log.

I’ll just quickly add that ASUS apparently has a “surge protection” that will display a message if it is triggered. Of course, I wouldn’t know if it’s overvoltage, overcurrent, etc, but I have never had that message show up.

12

I have considered this, but most of my USB devices are powered through a separate powered USB hub, where I can switch off devices that I don’t want to connect.
But I will try unplugging the hub entirely to see if that is the problem.

1 Like
13

Uncheck automatically restart and save changes.

image

14

Dumb idea: Plug the minimum number of usb’s needed to fly directly in to the pc, thereby skipping the usb hub. Just as a test. Good luck.

15

did you tried already limit the system load, eg. simple limit max fps ? You have then also reboots ?

But also search the issue on other sides… As example:

related to this there are existing topics and I remember some issues with USB devices/settings ( usb energy mode , hub , etc )… So far I know it caused only crashs (not reboots), but may be it is a sympthom ( less power for devices ).

Also… recheck all connections… we had some cases where user found a not correct plugged cable :slight_smile:

16

All 4 slots populated? What’s their operating temp? (HWInfo will show you)

1000w of PSU should be enough for that rig, the Ryzens are pretty power efficient aren’t they? But I recently had to swap out an old faithful BeQuiet 1200w that’s served me well for years because it was starting to get a bit flaky (sometimes refusing to start up, sure sign of failing PSU)
I swapped in a Corsair HX1000, not a high end PSU but it can handle a 3090rtx at full chat + a lot of other stuff. (8086k @4.8, 64gb ddr4 - 4 sticks, 2 pumps, 12 fans…)

I had to juggle some cooling solutions around when I put the ram in. 2 slots runs ok cool, 4 slots gets pretty hot…

17

It does run quite warm. I don’t remember the exact temperature, so I will have to check. But I don’t think it is overly hot, or else the temperature would’ve caught my eye, I think.

My PSU has age on its side (it’s only 5 months old), but of course, I’m still trying my hardest to check if it’s the root of the problem.

18

I have tried limiting max fps, but I don’t believe I have enough fps to reach the limit anyway. Regardless, I do see how that would change system resource allocations, so I will have to double-check that.

The USB suggestion has come up a couple of times, so I will look into that next.

If by “cables” you mean the power cables, I have rechecked both ends (the PSU end and the motherboard/GPU end). They’re all plugged correctly, and the stress tests show the 12V, 5V, and 3.3V rails delivering within spec voltages.

19

Now, THIS is something I have not looked into.

I guess now a system failure would just seize the computer or shutdown, but then at least I would know if it’s RAM (seize), or PSU (shutdown).

20

Exactly my plan as well. Just need a bit of time.

I need to unplug:

  • VR
  • The USB hub with all the yokes and sticks.
  • Speaker

and just replug the yoke with no rudder, and the mouse and keyboard.

If THAT also crashes, I think it’s then safe to assume that USB isn’t the problem.