PC is about a year old. Specs are Ryzen 9 5950x and RTX 3090, 32 GB of memory, 1000w PSU. It remains on and is useable for hours at a time, even under heavy load, but will randomly lose power amd restart. Event viewer claims it’s generic kernel power loss, which leads me to believe it’s a power supply issue. My PSU is a reputable 80+ Platinum model though, so I am hesitant to pull the trigger on buying a new one unless I am ABSOLUTELY sure that it is faulty. Obviously hardware of any quality can fail, but I do want to cover all my bases before dropping $130 on a replacement. Here is a list of everything I have tried. Is there anything else I should test?
42 Comments
I agree that it’s your PSU. What’s your warranty look like?
You could check the cables from the power switch to mainboard if not already done.
I have seen behaviour like this which seem to be caused by a loose cable.
From personal experience, that happened to me because of a faulty PSU.
My outlets aren’t grounded except in my kitchen, so the PSU might have started malfunctioning because of a blown up capacitor or something like that.
Check your PSU and make sure your outlets are grounded.
100% a psu issue. Either something is tripping over current protection or the psu itself is faulty. Try to warranty it
My PC would restart Randomly a few imes a week before i found out it was a bad stick of ram. Run mem test
[https://www.youtube.com/watch?v=PLr_sYeH30k](https://www.youtube.com/watch?v=PLr_sYeH30k)
Can you open windows event viewer after it happen and post the exact errors you have?
Just curious
I had the same problem, turn off RAM overclock in BIOS, this should resolve the issue
Do u use curve optimizer sometimes it’s unstable
I have same issue. I get kernal power 41 critical ID 63 and Bug Check (minidump) gpu driver issues. Try changing older drivers for your gpu. I took my pc to 3 diffrent stores, they couldn’t find anything.
Have you checked c-state settings in your bios?i wound up disabling c-states entirely. I was having similar issues and this fixed it for me. sorry if it doesnt help.
I’m having the same issue. Random reboots. Sometimes 3 in a day. Sometimes 1 in a week. Strange thing is it has never happened while gaming. Only while idle or web browsing. I’m about to buy a new PSU and see if that fixes it.
I encountered the same random restarts while gaming, after updating Nvidia drivers.
With a fresh Windows install, the issue was still there.
When I disabled the xmp the restarts didn’t happen.
I now run xmp with a manual clockspeed of 3400 MHz instead of 3600 MHz, and no restart happend since then.
I had a simular issue with a system with almost the same specs. Check so you run multiple cables from the psu to the gpu.
My problem was that the gpu was power starved and then faulted and that shutdown the pc and when the gpu tried to pull to much power it triggered over current protection
I had a very similar issue, and randomly found the fix on an old reddit post with the same problem. The solution was to increase ram voltage by .5 millivolts and amazingly it worked and the system has been stable ever since. Before trying that I had replaced the PSU, undervolted everything but nothing worked until that tiny power bump and now it’s perfect.
try a different PSU
note that if its modular and of the same brand you don’t need to unplug anything from the motherboard.
if its not the same brand you need to use them to connect to the MB as well
random poweroff without a bsod sounds like a PSU
the PSU could be at the point of it’s life where it can’t deal with a transient load spike, maybe it can deal with 800W sustained and not flinch but going from 120W to 400W in 0.1s makes it fall over itself
Ok this same thing was happening. I have a UPS, I moved outlets and it still was happening. I thought my GPU was overheating. Repasted and everythings fine. No more Random shut downs. Then GPU died randomly one morning. PC won’t turn on. I’m now on my old 960 lol. 5700xt died, my PSU is a Corsair SFX 750.
TLDR: Might be your GPU, similar events played out before it died.
I know you said your temps were okay. However, I had a very similar problem: random rebooting/crashing with no BSOD. It turned out that I had too small of a CPU AIO cooler. When I upgraded my CPU I didn’t upgrade my AIO. This caused my problem.
Once I upgraded my AIO to a proper one for the CPU everything worked just fine.
Curious if Event Viewer has any logs other than “did not shut down properly”
But as others said, probably a PSU. Try for an advanced RMA with replacement. Otherwise, I’d just eat the cost of the PSU if its something needed for work.
Be sure to run Memtest86/+ (I forget which is the better one in 2023)
Is the 32 GB of Ram 2×16 or 4×8?
If you go into your event viewer after a restart it will give you an error code and then it might point you in the right direction. But I do agree it sounds like a PSU issue.
I know you said you used a different outlet but are you sure that that outlet is not on the same circuit?
Recently had an electrician come and move some things around the circuit breaker box, now my PC reboots randomly.
Yup, your PSU
Is your PSU from Seasonic? had this happen to me twice, wtih focus 750w and gx-850 platinum, would shutdown pc if i was playing anything that made my evga 3080 pull power.
Seems like those 2 models had OCP set to trigger way to soon, and with the nvidia 3000´s transient spikes, that made the pc shutdown with games. But that did not happen on the first few months i had those PSUs, seems like that something that grows overtime to the point that even when loading windows will make it shutdown, idk, but after i replaced the 850w plat for a 850w gold corsair, the problem was permanetly fixed.
Have seen some suggestions to do a memtest, I would agree. Remember diagnosing / troubleshooting is also about weeding out all the false flags. May not be memory, but that will always be a possibility until you do memtest. PSU seems to be the most likely culprit, I worked at a local PC repair shop and we had 24 pin PSU testers. If I was feeling generous I could do it at the front counter for the person. Get the PSU tested and run memtest and I think you will be a lot closer to figuring out what the issue is.
Prolly PSU, but also had this problem with a faulty CPU once.
you can check the windows event log, look for a “critical power” or something like it, that would be a good indication that is you PSU is maybe faulty
do you have an update that’s waiting to install?
Installed Windows twice? Lol
Trying taking out ram sticks…mine was down to one dodgy stick which triggered a reboot every so often. No issues since replacing
Could be brown-outs (when the lights dim, but don’t go off completely). Get yourself a good sine-wave UPS.
If it still happens, it’s probably your PSU.
“typical scenario”
I had the same issue for YEARS.
It was the PSU for me, but it was not because it was faulty. Apparently, the 3000+ series cards sometimes have very minor power usage spikes. PSUs from before they were released can’t handle that and will randomly shut down the computer instead when a spike is detected.
Check when your PSU came out and if it supports the spikes the cards can cause. If it is an older model – change it out.
Ram. Its ram
from my experience this happened because my new CPU and GPU both were dying, had to rma
I’m not too knowledgeable when it comes to troubleshooting but when I hear strange PC issues such as this, RAM is strangely the common cause to these. No clue why though
I had this happen to me, make sure your bios is fully updated
Hello i had kernel power error and black screens or restarts back in the day and the problem was ”fast start up” thing in windows power settings – the ”choose what the power buttons do ” tab
Try a bios update. I’ve had it fix this issue before I’m not sure why but it did
I had a similar issue. Turns out my memory was bad which corrupted my drives(except my problem went farther and I actually had a bad CPU too, wild.) If the event logger says kernel power loss because of the shutdown I wouldn’t necessarily assume its the PSU. Any crash or hard restart will trigger a kernal power in the event logger.
I had the same issue, turned out it was the evga power supply I had. I switched back to a Corsair psu and haven’t had any issues since.
Bad PSU. I would first check to make sure all the cables are both plugged into fully to components and to the PSU. I’ve gotten burned by that before, do that before buying a new PSU