The PC turns off during games, yet the RGB remains active?
The PC turns off during games, yet the RGB remains active?
Here are some thoughts you might consider...
You could look into actual power consumption data, since spikes in usage might be causing system instability (this is often called a 'transient surge'). The PSU's wattage should match the GPU's needs, but it might not be providing enough.
It’s also worth reviewing the details in your motherboard manual to confirm there are no restrictions on slot sizes or speeds—some slots share bandwidth with specific M.2 ports or SATA connections, which can affect performance when running at high speeds. Also, whether this applies to the RX 7800 XT isn’t clear, but from what I understand, certain AMD GPUs really dislike PCIe3.0 speeds. You might want to check your motherboard’s BIOS settings; some allow individual PCIe slot speeds to be capped at 3.0 instead of 4.0, which could impact performance if set that way.
Although this isn’t a Ryzen build, I’ve noticed that RAM can be a big issue for Ryzen systems. Following the board manufacturer’s QVL (Qualified Vendor List) is highly recommended for stability with Ryzen, as it ensures compatibility with memory and storage. An Intel setup could face similar limitations.
Perhaps the first sign is that your motherboard’s QVL doesn’t include any 3600 DIMMs from any manufacturer. The slowest ones listed are 4000, while the slowest TeamGroup kits are 4800. It seems your system might be struggling with this limitation. These slower DIMMs could lack the speed needed for stability, and it’s surprising the board runs at all.
For context, I’ve seen several systems encounter problems with certain DIMMs that weren’t listed in the manufacturer’s QVL—switching to approved kits resolved the issues completely, with no errors in tests like memtest86.
BIOS behavior can vary; some systems showed inconsistent speeds depending on the speed setting.
Shutting down during gameplay with fans running or off, sometimes with RGB lighting, occasionally without.
Not waking up from sleep.
No video output after a cold boot (8–10 cycles), possibly due to memory issues.
A constant hourglass cursor in Windows.
I hope this gives you some insight.
Fair comment. I'm the only PC enthusiast in my friend circle, at least locally. I'll check what I can do.
I'll run DDU and reinstall my drivers.
TLDR: I really don't remember the forum. I was away for two days after posting to mark the 4th, and when I returned I got an email with two replies—one asking for more (and calling me cheap for buying a B-Tier PSU) and another saying the thread was closed because I didn't respond. When trying to log in, the site kept redirecting me back to Google home. I'd tell you about it, but I deleted the email since checking it in my Gmail made me annoyed.
Ran DDU. It's now 3 AM, so I'm not in the mood for gaming. I'll update you tomorrow on any further issues. Thanks.
Many complicated terms here, right?
Do you think a surge protector could stop sudden spikes? I have my PC connected to one of these.
You might also want to read the small details in your motherboard manual to confirm there’s no problem with the x16 slot being replaced by something slower (like some M.2 slots or SATA ports). Also, it’s unclear if this rule applies to the RX 7800 XT—some AMD GPUs really don’t like PCIe3.0 speeds. You could look into your motherboard’s BIOS settings; they sometimes let you set individual PCIe slot speeds lower than full speed.
I went through my manual and didn’t find anything about this, but I might be missing something. I’ll double-check the BIOS to see if the PCIe slots are limited to 3.0 instead of 4.0.
Just a note: my memory is in the supported list. From the link you shared, it seems you’re checking the DDR5 version, but I have DDR4. My games SSD was listed, my boot SSD had a higher capacity version, and the extra storage SSD wasn’t included.
Just to clarify, my PC has been working well for months, but this problem started only recently (sorry if it sounds rude).
Any thoughts you have?
Strangely, mine ran fine for a while before it started acting up too.
Not a dumb question for someone who isn't experienced -- in fact, my old man told me the only dumb question is one you never ask, because you'll never know the answer.
But no, a surge protector strip has nothing to do with consistent power from your PSU to your other components. Transient surges are basically momentary spikes in CPU / GPU power demand, resulting in a PSU overload in some cases, which will cause the PSU to shut down to protect itself, just like tripping a circuit breaker, and will causing the system to crash from power loss.
This can occur even if your PSU meets the required power rating -- basically, if your PSU is 'adequate', it could still overload from transient surges. This is why I chose an 80+ certified 850W for my 5900X / RTX3060ti rig. But even if you exceed the requirements, there's always the possibility that your PSU simply isn't delivering the wattage it is rated for.
Interesting that there are two versions of this board. Is there a link for the one you have?
As for the PCIe slot limitations, in plain English, you should read the manual sections pertaining to the PCIe x16 slot (the longest of them, and if there are multiples of the longest length, the closest to the CPU), the M.2 slots, and the SATA ports.
There will be footnotes indicating if any specific configurations will cut speeds, disable SATA ports, etc., and there is no particular standard to this -- it varies from manufacturer to manufacturer, and board model to board model. There may well be something choking your PCIex16 from 5.0 to 4.0 or even 3.0, or from x16 to x8 speeds.
As an example, here is
such an excerpt
from the manual for my Gigabyte B550 Aorus Master.
Hope this helps make more sense of what I'm saying.
Hello. Just checking in here. You happen to have the exact same issue I am having. I was just about to create my own post before I saw yours.
My issue is identical. PC shuts down, fans turn off, but some internal RGB remains on, indicating there is still some power flowing. PC cannot be shut off / restarted except by turning it off at the wall. PC is four years old and only started experiencing these issues this year.
Just to highlight a few things I've tried - I've run memtest multiple times. No issues discovered.
I've attempted to trigger a crash via CPU and GPU load tests. This has never worked.
I have checked the system journal (I primarily run Linux) for any offending errors. Found nothing.
I even brought a cheap graphics card and replaced my existing one temporarily. Crash still occurred.
I've reinstalled drivers in both Windows and Linux (dual boot). Issue still occurs across both OS's, even though both are on entirely separate storage devices.
At this point I've ruled out the graphics card, the RAM, the storage drives, drivers and operating systems as being a potential cause. The fact that CPU load tests never seemed to trigger a crash makes the CPU look innocent (although I can't be 100% certain). The two most suspicious looking suspects remaining are the motherboard and, like you said, the power supply. Both are a few years old, so wear and tear is a factor.
My motherboard is the ASUS TUF Gaming B550M Micro-ATX board.
My PSU is the SilverStone Strider ST65GS.
Do you know what motherboard and PSU you're using? If we have one in common, that might provide a clue as to the culprit.
Honestly, I've noticed PSUs often come in a mixed bag, created by various manufacturers. PSU is just a brand here.
I've used seasonic focus line or core gold line PSUs without any issues.
In reviews, I focus on common brands like Seasonic, Great Wall, FSP, and others.
They generally make decent PSUs, though some from FSP have been problematic.
It's best to stick with single rail PSUs.
Also, didn't you mention the type of RGB—whether it's strip or fans? That matters because different RGB options use 5V vs 12V and similar, and plugging in the wrong one can damage the board.
Curious... mines 850W 80 Plus Gold and still might be acting up.
Here's a
link
to my board. Mines specifically rev 1.3
I checked the
manual
like you said and see nothing that suggests any bandwidth sharing.
That does help, thank you!!!
I finished playing about an hour of Overwatch 2 following Lutfij's suggestion of using DDU to reinstalling my drivers and so far so good. Then again I was running stock without overclocking. I'll do some more testing now...
Apologies. It's ARGB in the fans, RAM, motherboard, CPU cooler, and GPU. As mentioned before, everything is good so far, but I need to conduct more tests, particularly for overclocking. I bought the entire GPU and PSU, I really want to use the whole GPU and PSU!
Overclocking usually raises power consumption, so combining GPU and CPU overclocking might push the system into temporary spikes if demand matches supply. For instance, my setup uses 175W when idle but can spike to 280W during intense gaming. It’s surprising how little guidance there is in that manual—just jump straight into BIOS settings. It reminds me of Pee-Wee Herman shouting, “LA-LA-LA! Connect the dots!” I’d recommend testing a power supply tester if possible. If you’re near Orlando, FL, Greg Salazar’s YouTube channel “Fix or Flop” might help—he has the right tools to troubleshoot persistent issues.