System question about restart?
System question about restart?
my system has been experiencing frequent random restarts without apparent cause, and I’m unable to reproduce them consistently.
Here’s my setup:
cpu: ryzen 5 7500F (undervolted with PBO all cores negative 20)
ram: lexar thor ddr5 6000 cl32 (expo profile)
gpu: xfx rx 6950XT (multiple profiles; recent crashes occurred on this one)
min freq: 2200 MHz
max freq: 2300MHz
vram: stock
fans: 80%
power limit: stock)
mobo: asrock b650 pro rs
psu: 750 W gigabyte UD
I also noticed several types of restarts and reboots.
1. The screen turns completely black or green with a digital buzzing sound for a brief moment before restarting. I believe this is due to the vram overheating or being overclocked, as I can reliably recreate the problem under similar conditions.
2. The system restarts abruptly without warning, and the Windows Reliability Manager reports something like "the system shutdown was unexpected." I can’t reproduce this issue, but here are some additional details: it happens regardless of load or otherwise, sometimes after a stress test or while streaming on Steam.
I previously undervolted the cpu and then reverted to normal settings after running stress tests (OCCT, Prime95 for 13 hours). I also adjusted ram timings and performed stress tests with OCCT, Prime95, Memtest, and everything functioned correctly.
After some attempts, I finally realized the issue might be related to a restart/reset button or connection. I unplugged it.
If you need more details, please let me know. This problem is really challenging to troubleshoot.
Thanks in advance.
I examined the PSU and noticed that while you might be correct initially, the frequent PC reboots without any activity raise doubts about the "PSU is a problem" argument.
My power consumption stays below 80W for the CPU and 310W for the GPU.
It's improbable that all system components would suddenly spike together to surpass these limits.
Here’s a good video to check out: View: https://www.youtube.com/watch?v=wnRyyCsuHFQ At around the 9:40 point, he begins discussing transient spikes. The power consumption for your card is listed at 335 watts, which doubles during the transient spike to 670 watts, plus an additional 80 watts for the CPU. When you add in any other system components, it reaches 720 watts overall. It only requires running a game and then closing it—this causes the card to drop out of load because there’s nothing to render. Suddenly, I need to boot up the desktop, and the transient spike appears. It happens in just a split second, and most software doesn’t refresh quickly enough to detect it. I’m sure you’ll understand why you have a 1600-watt PSU installed. My original EVGA 750-watt G2 model failed under transient spikes from my 2080 Ti FTW3. Rather than purchasing a 1000-watt PSU, I upgraded to an EVGA 1600-watt G2 that I had previously saved for another build. If you’re looking closely, here are some details: - GPU: 670 watts with transient spike - CPU: 80 watts (as mentioned) - NVMe drive: 10 watts during read/write - RAM: 3 watts per 8GB, assuming 16GB at 6 watts each - Fans: approximately 4 units total (case and CPU cooler), about 12 watts That adds up to around 788 watts in an ideal setup. Also consider the age and degradation of your PSU. If you’re using an AIO, expect around 25-35 watts. For more details, check out this review: One of the noted drawbacks was “Transient response could be better.”
I would suggest it's the power supply unit as well. It isn't a good choice for handling sudden voltage spikes, which leads to earlier wear from those surges. Since there hasn't been any BSOD reported, it's less likely to be a software or driver problem. The spikes occur extremely quickly, making it hard for the software to notice. A simple test would involve running your system with an 850W PSU that's known to work properly.
Unfortunately I don't have another PSU either, and I can't get a new one anytime soon. Also, a PSU failure to respond to sudden loads might cause a system shutdown instead of just restarting. Sorry for asking so many questions—I'm trying to learn and don't have much PC experience yet. But I've undervolted and underclocked the GPU to the point it only draws up to 230 W, while the CPU uses about 80 W at peak, which is unlikely to be a game. Still, I'll keep testing. Thanks for the replies. Edit: I should have mentioned the PSU is only a few months old.
It's worth mentioning that the gap between GPU temperature and GPU hotspot typically ranges from 15 to 20 degrees, with extreme situations reaching up to 30 degrees.
Restarting after power loss could stem from BIOS power settings. Setting POWER ON following AC power loss might account for the issue. To keep the PC off, check for options like 'AC power loss restart' and disable them. Spikes occur suddenly, briefly doubling the load beyond normal levels during gaming or stress tests (the PSU rating). Often, the PSU’s OCP triggers a shutdown due to these spikes. This may reduce wear from age if the unit is new. Based on your situation, it seems likely the device struggles with spikes. If you notice this more often in specific games, avoid them temporarily until a more powerful unit arrives.
I incorrectly assumed the PSU wasn't the issue. The definitive proof would come from installing a verified PSU. Simply restarting without action doesn't confirm the power supply isn't the cause.