PC Restarting itself at random intervals during gameplay

User Support Hub Hardware & Peripherals
1

Hello All,

First of all, i just want to say that i Have googled this issue extensively, and i have read many similar issue to my own that relate to a under-powered/faulty PSU. i do not believe this is the case in my situation, and i will explain why below.

My PC Specs,

CPU: AMD Ryzen 3900X
CPU Cooler: Corsair H100 360mm
Motherboard: Asus Prime X570 Pro (Latest Bios)
RAM: 32GB [2x16gb kit] G.Skill TridentZ Neo (4x8gb)
HDD: [OS on 1, Games on 2] 2x ADATA m.2 NVME SSD (Gen 3)
GPU: Gigabyte Nvidia 3090 VISION OC 24GB DDR6X
PSU: Corsair RM750x 750 Watt
Display: 2x Acer Predator XB273KGP 4K 27"
Controls: Alpha Honeycomb Yoke ,Alpha Honeycomb Throttle Quadrant, Logitech Rudder

Just to clarify, my average Wattage pulled from the wall is 600W during peak gameplay in Cyberpunk. FS2020 Does not utilise the CPU efficiently so the wattage pulled is closer to 550w.

The Issue Symptoms:
At random intervals while flying in FS2020 my PC will simply Reboot itself without any BSOD or errors.
BSOD Viewer reports no BSOD since the OS install. I installed the OS 2 months ago after a hardware change. (Moved from Intel to AMD)

I have monitored my CPU Temperature. It averages between 50-60c during FS2020 gameplay.
The GPU averages around 55c during gameplay, with a Memory t-Junction temp of around 68-75 degrees.

What i have tried.

  1. Swapped the PSU with an Corsair RM 850X 850w.
  2. Lowered the GPU Power limit to 70% (250w) from its average 350-360w at 100%

In both of these instances, there is no change and at random intervals, It could be after 15 minuets, or 3 hours, my PC will look like its crashing to the desktop, and then the PC just reboots before i get to see the desktop.

Note, on my 2nd 4K display i am usually watching Youtube, or have OnAir Company running but no other heavy loading software.

In some instances, But not all, The reboot can happen when i have moved my mouse to my other display and when moving it back to FS2020 and try to rotate the camera, this can cause the PC reboot.

I have reached the point where i don’t want to play the game anymore because this issue is really frustrating as i lose my progress in my planned flights. I am mainly using the KingAir 350.

I hope someone can recommend some solutions or advice, because im at a deadend with my research into this.

Thank you all!

1 Like
2

Only with MSFS or any other intensive programs ?
Your Mem/CPU/GPU look ok.
Peripherics drivers ?

3

If you believe it’s not the PSU, you could investigate separately the CPU, GPU and RAM with OCCT.
I like this tool because it’s simple and effective. According to my experience, if I have a green tick after an hour of test it means that the component is rock stable.

4

I will give it a try,

Yes it only happens with MSFS so far, i have updated mouse/keyboard/all motherboard drivers.

1 Like
5

Look for entries in event viewer around the time the reboot happens. There might not be anything useful, other than a warning about an improper shutdown but its worth a shot.

1 Like
6

The only time this happened to me was when I attempted to overclock the CPU or GPU. If you haven’t overclocked anything, try to downclock your GPU since it appears to be stock OC…
If that helps, then RMA the GPU

Also, keep in mind that lowering the GPU power will make it less stable at higher speeds

1 Like
7

gut feeling on reading the above which is a very nice rig is power , the nvidia stated limit for a founders edition is 650 watts whereas aftermarket cards can draw even more power with 3 x 8 pin connectors and your running a respected brands 750 watter but it feels a little low in juice for such a well specced machine , case fans as well can make an impact with their draw as your machine is keeping nice and cool so cooling is working well at those temps but that also enables a card like the 3090 to run higher clocks with its power/clock curves deciding how fast it can run at given temperatures which then leads to it drawing more power when it sees it is keeping cool so in theory it could demand a power increase near to max by virtue of the decent cooling

possibly try a 1000 watt psu , no bsod or error reports point as well towards instantaneous shutdowns which a leading cause is power supplies even corsairs , but also try swapping memory sockets around if your 100% certain on the power , other possibles … swap your usb connections around as this usb disconnect on amd systems may have a hand in your problem

8

Remember that someone had the same issue when Asus increased the default VDDCR SOC voltage to 1.1v which causes the system to reboot randomly. https://www.reddit.com/r/ASUS/comments/kh1hjx/prime_x570pro_random_reboots_on_bios_3001/
Maybe worth checking…

1 Like
9

Just an update,

Fistly to respond to Mushioni, taht was my initial thought, but after running several run thought PSU Checker incloding adding LED’s and fans and AIO pump etc, it still runs under total 650w , and the gguidense from that site recommended a 750w fot the 80% rule. as mentioned, i have ran with a RM850x and the issue persists. i dont have/afford a 1000w PSU just to test this unfortunately.

Upon reviewing the system event log. Just before my crashes i see this error.

“The description for Event ID 14 from source nvlddmkm cannot be found.”

image
image2419Ă—967 129 KB

image
image2433Ă—398 47.7 KB

it seams to me this is a common problem among other users of AMD and Nvidia

1 Like
10

a lot of users are pointing to power plans with this crash as well , as you say the psu looks ok but possible windows power plans may not be working for you , i installed ryzen master and followed this and use the high performance ryzen plan on my own

ryzen balanced power plan

1 Like
11

Are your system overclocked? Try it totally “clean”.
Check the temperature from your GPU during gameplay…some guys had this problem, and solved making a better airflow to avoid PC shutting down as a protection due the heat.

12

I second this I agree with this I think the OP should be going with 1000w with all of that stuff he mentioned.He could be indeed short on juice.

13

It’s possible the GPU is causing the loss of power, going by the error messages. It may still be a PSU issue. I have a 1600W EVGA PSU, and have never had this problem.

14

it’s not really the average sustained load you should be worried about. It’s the really tiny spikes in wattage that can trip a system up. A 750W PSU is on the lower edge for a 3090, especially if it’s a bronze or silver rated one (gold or platinum should be ok-ish).

1 Like
15

How hot does your memory run?

4 sticks means it’s not got much space between, especially with those wide Trident-Z RGB spreaders.

Increased heat = increased chance of error.
It only takes one error in memory to crash a system.

16

No i run stock.

17

Update

Ran Stability tests on CPU/GPU all ran without insurance.

Memory Temps are between 30-40 degrees.

CPU Temp 60c
GPU 65c

I have resolved the issue, i know a lot of you say, I have 1600w PSU and i dont have that issue. or your PSU cant handle the PEak performance. but this is sadly not correct in this instance and 750w is capable of running a 3900x and a 3090 stock without issue.

The solution to my problem was simply toggling a setting in the Nvidia driver.

image
image863Ă—734 37.1 KB

Setting this to Prefer maximum performance stops FS2020 from reducing and increasing the gpu power randomly.

This could be an issue with the AMD power profile.

Hope this helps someone else who struggles with this, and dont just buy a $300 PSU when not needed.

5 Likes
18

it ties in with the power plans in some way , nicely done now enjoy that beast of a rig :clap:

1 Like