General Discussions

[LaCarognaSulleSpalle]

Fresh build, here are the specs.

Mobo MSI B650 GAMING PLUS WIFI (Updated to latest firmware 05272024)
CPU AMD Ryzen 7600x
GPU XFX RX 6750 XT Speedster MERC 319
RAM DDR5 32 GB (2 x 16 GB) Zenith 6000 MHz (PC5-48000) 288 pin CL30
PSU CORSAIR RM650 ATX 650W (modular)
Cooler Thermalright PS120SE ARGB
OS Windows 11 Pro
SSD LEXAR SSD M.2 1TB PCIe Gen 4X4 NM790 NVMe
HDD Seagate Barracuda 2TB SATA III

I am pretty sure this must be a GPU-related issue, since the system wors very well with every other task. I found that I can play with my console emulators just fine, which makes sense since emulation relies mostly on CPU. However, every PC game I try to launch doesn't last more than a few minutes, before crashing. UE4 games give me a Fatal Error with no specification or error code whatsoever, but sometimes there's not even an error message, I just move my eyes and the moment I look at the screen again I just see my desktop.

Sometimes, it even crashes while installing AMD drivers and when the system reboots, I get amdnoisesuppression.exe bad image error. I think, though, that the first time I ended up with that error, I could play for a longer time with no issues — a longer time being 10 minutes, mind you, because that error message bothered me and so I uninstalled and reinstalled the drivers again and fell again into the cycle.

Before posting here, I googled everything I could.

• I've been through a Sisyphus' cycle of uninstalling (via DDU in safe mode) and reinstalling every possible version of the drivers, with Full Installation, Minimal Installation and drivers only. I tried the latest drivers, then I tried the older ones that were said to be more stable, than I even tried random ones.
• I disabled overclock, game boost, XMP from BIOS
• I also disabled MPO
• Chipset and bios are up to date.
• I set balanced mode and Max Processor State to 99% in the power mode settings.
• I made sure that the RAMs are placed in the correct slots
• I tried to make sure that the GPU is well inserted in its slot as well.
I've tried checking Windows Event Viewer and Affidability Report, but either I couldn't find anything relevant or didn't know what to make of it.
I think I have shared any piece of info I could, but please let me know if you need any more details.

I think I can also hear a soft buzzing sound when my GPU is active, i.e. when I launch a game. I really can't tell if it's just a sign it's active and working or it's a symptom of something wrong.

EDIT now, OCCT gave no errors. Doom 2016 runs smooth as butter. Hellblade hasn't crushed yet (switched to DX11 though). Lol. What the funk. Seriously. I'd delete the whole post but just to be safe I'll wait a couple days.

EDIT JUST GOT ANOTHER CRASH ON SUNSET OVERDRIVE. SERIOUSLY. WHAT. THE. HELL. WHAT. THE. ACTUAL. HELL.
EDIT tried memtest86, 1 single error on test 9 during first pass, none in the following ones.

[msr062006]

did you ever find a solution?? i have the same mobo and im having the same crashing problem, please help if you can thanks!

[LaCarognaSulleSpalle]

Hi! Please check my answer to elstaci's reply, it might give you some hint.

[elstaci]

By any chance did you run OCCT PSU Test?  Did it pass without crashing and restarting?

Did you check your PSU Outputs during the testing (3.3/5.0/12 VDC) all within 5% tolerances?

It is possible your 650 Watt PSU might not be able to handle very heavy loads.

According to AMD you need a minimum PSU of 600 Watts with your GPU card but with your 7600X it might be pushing your PSU past its output limits.

This CPU/GPU Calculator mentions you need a minimum 600 Watt PSU for your processor and GPU: https://www.whatpsu.com/psu/cpu/AMD-Ryzen-7600X/gpu/AMD-Radeon-RX-6750-XT

The above website PSU requirement are normally much lower than the actual Minimum PSU Requirements. Seems to show the actual power usage but doesn't seem to include Power Transient voltages that could crash a PSU.

If you PC just shuts down by itself it could be a power issue with the PSU under very heavy loads like gaming.

Also Overheating can cause a PC to shut down without any errors.

Poorly done Overclocking/Underclocking can do the same thing.

Defective or incompatible hardware can do the same thing.

Note: on Memtest 86 you shouldn't get any RAM errors at all. It should all pass with zero errors.  Run the test again with just one stick of RAM installed and see which stick is the one that gave you that RAM memory error.

[LaCarognaSulleSpalle]

Hi! Thank you for the hint, I didn't expect to get any reply after more than one month. I'll tell you how the situation has evolved.

I have looked for advice on reddit and somehow ended up fixing my main problem (games crashing) by overclocking the CPU to 5.2Ghz. and enabling XMP for my RAM - I don't even remember how I set it, I was just fumbling in desperate and desperate to make it work no matter what. It worked and now I can game without any problem.

HOWEVER. There is still ONE operation that causes my CPU to completely restart - video rendering and/or encoding. So I tried OCCT PSU test, as you suggested. My PC crashed and restarted the moment I started the test.

So, I guess this suggest the culprit of all problems is my PSU isn't it? Should I change it? I'm afraid it's a bit too late to return my current one to Amazon...

[elstaci]

What OCCT PSU test does is run both CPU and GPU Stress test at the same time putting maximum load on the PSU.

In the past, if you had a weak or poor quality PSU the test might end up damaging the PSU but it won't damage High Quality or strong PSUs.

Thanks for the update.

[BigAl01]

Replace the PSU with a quality (name-brand) unit.  

[Mooncatt]

I'm having an almost identical issue and I think I'll try this now too (though I can run Cinebench R23 and Heaven Benchmark at the same time with no issues).

I'm using a 5800x3d and XFX 7800XT on a Corsair RM750x, only this is a recent problem for me. I was running rock solid until the GPU driver updated to 24.6.1. That's when I started getting instant crashes and reboots with any game, which still exist on 24.7.1. If I roll back to 24.5.1, everything is rock solid again.

You said this is a new build, so have you tried an earlier driver yet? I would if you haven't already, and I'll try to run mine through OCCT and maybe we can compare notes. I started my own thread about it yesterday if you want to go look at it. I only just now saw this one.

It would also be interesting to know what happens if you try to under clock and/reduce the power limit of the GPU. I did both through Adrenalin with no change 

[BigAl01]

If you were running fine with the current PSU and then it started crashing after a GPU driver update, I would not replace the PSU.  There is something else causing your problems.

[Mooncatt]

Which is kind of my thought and also why I suggested trying an older driver if this is a new build that hasn't previously ran on one. If I were to have built mine last week brand new, I would have jumped right on 24.7.1 and having these issues from the start. That would more directly lead to a PSU concern when it probably isn't. 

One possibility I was pondering was maybe something in the driver was allowing for a massive power spike in certain hardware. I think we saw a similar situation with the New World game that was killing GPUs. I did just try running the OCCT Power test with an interesting result. It ran fine on 

.5, but gave me the same shutdown/reboot the first time I tried to launch on .7. By that, I mean it didn't even fully load OCCT before it crashed, not even making it to a test to start. It was only that one time, and runs fine on .7 ever since. I'm now letting it run the full power test while I write this, but a positive sign that it didn't immediately crash on the test start.

I'm will to accept that my issue is different, but it was so similar that I wanted to chime in just in case there is something related between the OP and me.

[BigAl01]

It sounds like a PSU issue that cannot handle the load.  I only build computers with 1K watt and higher PSU's (name brand too) to ensure they don't cause me issues.

[TwinTurboRYZEN_PLUS-ULTRA]

I was getting random reboots that didn't leave any type of log until I looked into my PSU.  It has an overcurrent protection that will just shutdown and reset the PSU. 

It was a PITA to figure out the cause of that issue since it would only happen with certain games so it wasn't easily repeatable.

[elstaci]

Good reply but why add a SPAM link at the end of your reply  to a website that has nothing to do with this thread.

[BigAl01]

If this was happening to me, I would replace the PSU with a name brand 1K Watt or greater unit.  I would not keep dealing with this until I was sure the PSU was not the cause.  Next up would be looking closer at the memory sticks.

This baby is 1200 Watts and will handle everything you can throw at it.  Currently, it's priced at $200 USD.  I have one and it's working just fine.

Your PSU only provides the power requested from your load. It will not be putting out 1200 watts unless your computer is demanding that.