Random black screen - GPU or PSU bricked?

[Provin915]

Hi all,

Built a new PC in August and everything was running fine until last couple of weeks. I randomly get black screens when I start a game or in the middle of a game. My PC stays on but I lose connection to my monitor. At first, I would turn off my PC wait 5 minutes and turn it on again. It worked. I sometimes would also unplug all cables and plug them in.

But nowadays, it is happening a couple of times a day AND when I restart my PC; Windows does not recognise my GPU anymore (unrecognisable in Task Manager). My resolution also turns to 800x600. I then have to go into device manager, disable and enable my GPU again until it is recognised.

I have the latest BIOS update / latest chipset and GPU drivers. I do not have any sketchy programs, I did not change a single hardware or software. It is weird, because in the first 3 month's I had zero issues. I could play any game, maxed out for hours but these last couple of weeks it just randomly happened.

Below my specs.

CPU: AMD Ryzen 7 5800X3D
CPU Cooler: Noctua NH-D15 Chromax.Black
GPU: Sapphire RX 7900 XT Vapor-X
RAM: G.Skill Trident Z Neo 16GB 3600Mhz
MOBO: Asrock X570 Steel Legend
SSD: Samsung 980 PRO 1TB
PSU: Corsair 750W RMx
CASE: NZXT H5 Flow Black
MONITOR: LG Ultragear 27GP850-B

Nearly every component is new, except the RAM / MOBO / PSU and Monitor. I ofcourse updated my MOBO for my CPU to work.

I have exhausted many different options, but it seems I still have this issue. My temps are also max 75-76 degrees Celsius under load (both CPU and GPU). I do use CUSTOM cables from Cablemod, I will remove these and use stock cables next time I get a black screen.

Could it be that my PSU is not enough? It is strange because the first month's there were no issues.

Could it be that my GPU is bricked? I had read in a couple of posts that the thermal paste was wrongfully applied on 7900XT. I have no idea what to test next.

Is my GPU damaged because of the tight fit of my CPU Cooler?

edit 1: RAM XMP profile activated running at 3600Mhz
edit 2: using default mode on AMD Adrenalin; turned off AMD FreeSync (someone suggested this on Reddit)
edit 3: seems to only happen when I start a game
edit 4: replaced custom cable with original PSU cable
edit 5: furmark test result https://gpuscore.top/furmark/show.php?id=1269960

 

[Macro Device]

Could it be that my PSU is not enough?

Only if it's defective. This PSU, given it works as intended, is more than enough for your system. What are the readings under heavy load?

Could it be that my GPU is bricked?

Yes. Get yourself a different GPU (100% known to work properly) and see if the issue persists. No = your GPU is toast, time for RMA. Yes = most probably it's on your CPU. Get any compatible CPU and see if it changes the outcome.

 

[Provin915]

Only if it's defective. This PSU, given it works as intended, is more than enough for your system. What are the readings under heavy load?

Yes. Get yourself a different GPU (100% known to work properly) and see if the issue persists. No = your GPU is toast, time for RMA. Yes = most probably it's on your CPU. Get any compatible CPU and see if it changes the outcome.

I just tested with Furmark (don't know if that's a proper test for checking PSU under load) but it says around 360W.

I just sold my older GPU can't really test atm.

 

[Macro Device]

don't know if that's a proper test for checking PSU under load

Nah, Furmark is a proper GPU roasting tool. Want your GPU dead, run Furmark for a couple hours. xD

Run OCCT (all-in) and post these voltage values.

I just sold my older GPU can't really test atm.

Getting a bollocks level GPU just for the sake of testing is easy and useful.

 

[Deleted member 229121]

It's likely the GPU - does this issue only happen when gaming? If not then it could also be the display.
I sincerely doubt it's the CPU - it would likely be something with the board before the CPU unless you're not running stable settings.

Using another GPU to be in known working condition is going to be the fastest way to sort this out.
Unless you have an iGPU on your chip you should make sure to keep a spare handy.

 

[Vya Domus]

At first, I would turn off my PC wait 5 minutes and turn it on again. It worked.

Why did you have to do that ? The PC wouldn't turn on after for a while or what exactly happened ?

Nah, Furmark is a proper GPU roasting tool. Want your GPU dead, run Furmark for a couple hours.

Used to now it's the complete opposite, it has become worthless as a stress test, playing fortnite for 5 minutes gets me higher temps than running furmark for half an hour, drivers detect it as a power virus and the card throttles more than usual.

 

[Nike_486DX]

yep, looks lke you get the typical error 43, which means time to rma your card. Even though i am not a fan of corsair-branded psus (they are known to be hit and miss) in your case its not the problem, as you mentioned that sometimes after rebooting you get error 43 or "basic display adapter" that definitely means your gpu is not being recognized properly, so time to rma.

 

[Provin915]

Sorry for necro'ing the topic all.

The problem was fixed. Apparently, one of the customised PCIE cables from Cablemod were damaged. They sent me a replacement set (X3 PCIE cables) and everything is working again no problems during gaming etc.

Topic can be closed! Ty all for the help, appreciate it!

It's likely the GPU - does this issue only happen when gaming? If not then it could also be the display.
I sincerely doubt it's the CPU - it would likely be something with the board before the CPU unless you're not running stable settings.

Using another GPU to be in known working condition is going to be the fastest way to sort this out.
Unless you have an iGPU on your chip you should make sure to keep a spare handy.

It happened mostly during gaming and sometimes even during idle time just browsing internet.

Why did you have to do that ? The PC wouldn't turn on after for a while or what exactly happened ?


Used to now it's the complete opposite, it has become worthless as a stress test, playing fortnite for 5 minutes gets me higher temps than running furmark for half an hour, drivers detect it as a power virus and the card throttles more than usual.

I figured I would turn off my PC and check the cables if everything was put in correctly, which it was.

yep, looks lke you get the typical error 43, which means time to rma your card. Even though i am not a fan of corsair-branded psus (they are known to be hit and miss) in your case its not the problem, as you mentioned that sometimes after rebooting you get error 43 or "basic display adapter" that definitely means your gpu is not being recognized properly, so time to rma.

Luckily, there is no problem with the GPU but the custom cables from Cablemod were somehow damaged I believe.

 

[FreedomEclipse]

I left a like just for the 'fat boi' GPU

 

[TechHands]

Seen this thread viewing the forum earlier today. I was going to start a new thread first I'll make a post here. Yesterday ago overclocking a system lead me into the troubling issue of a infinite looping power saving mode black screen that's right after the bios loading screen. Going into safe mode using DDU the system will boot fine then once I install the latest 24.4.1 amd Radeon driver suit the infinite looping power saving mode black screen come right back forcing me to enter safe mode to repeat DDU. When I first finished building the system GCC installed a newer driver suit that was 24.10something... now that I am in part business of troubleshooting an issue that download no longer is showing up in GCC so like whatever. My GPU is a rx7800xt sapphire nitro. Imma be troubleshooting this right after I post this reply EDIT just opened GCC and am5 graphic driver is now showing 50% until download completes soon I'll know if the issue is resolved if install of this is a complete finish. I came to make this edit to correct myself on the GPU driver version on GCC it is not a newer version actual version is 23.40.28.240312:I guess I had myself believing it was newer sense that once I went to install 24.4.1 a message prompt come before me the screen saying "did I want to downgrade" or something like that. Anyway peace take care

 

[NeoThuNder]

i had this issue before with my other pc build.
but for me its Motherboard. The issue is PC is turn on, then login to windows, and after 5 or 10 minutes, its going to dark ( Black Screen ). All hardware running like fan, rgb, etc. But i cannot do anything beside of restart manually by press turn on/off button in pc case.

In yours issue, its look like your GPU need to RMA.

If you have friend who willing to borrow his GPU, maybe you can try it first.

 

[TechHands]

I'm going to delete the above post, I looked over it looking at what I described and I honestly don't remember it I could had worked through the issue faster than i expected leaving it behind as fixed and resolved I can't remember any troubleshooting also. Ah well just wanted that to be known for anyone looking over things like the posts and all. Take care

 

[Bowser]

The problem was fixed. Apparently, one of the customised PCIE cables from Cablemod were damaged. They sent me a replacement set (X3 PCIE cables) and everything is working again no problems during gaming etc.

@Provin915
How did you determine that it was faulty cable? I am facing similar problem, but I cannot replace just a cable, I will probably need to replace whole PSU.