[SOLVED] PC turns off when stressing GPU

Aug 25, 2019
15
2
15
0
Hi Guys,

i just built a new system:

AMD Ryzen 7 3700X
MSI MPG X570 Gaming Plus
16GB G.Skill Trident Z DDR4-3600 DIMM CL15 Dual Kit
8GB Sapphire Radeon RX Vega 56 Pulse Aktiv PCIe 3.0 x16
Seasonic 650 FX Focus+ Gold

Windows 10 is fully upgraded.

But when I try to play a Game the Computer shuts off after like a minute.

I already lowered the cpu to 3.4 GHz and put the Ram to 3.2 GHz (CL 16).
But this didnt solve the issue.
To nail the problem down I used CPU-Z to stress CPU. -> Nothing happens ... runs stable
Then I used memTest (6 of them 2.5Gb each) to stress RAM. -> Nothing happens ... runs stable
Then I used FurMark 3D to stress GPU. -> System crashes after like 20-40 seconds.

EDIT: When i say system crashes I mean the computer turns instantly off (Blackscreen). Then the computer tries to reboot but usually needs several trys until it gets back into windows.

When the crashes occured the GPU had 60° C or 140° F according to WattMan.

I am really struggeling to find out whats the issue here guys.

My current thoughts are: Is the PSU to weak ? But 650 W should be okay right ?

Or is the Graphics card broken ? But how could i then just write this Post here on the Computer ? System is perfectly stable unless is stress GPU

Any advice would be really helpful.

Thanks in advance :)

Mario
 

PC Tailor

Distinguished
Herald
Welcome to the forums my friend!

As a first port of call, have you updated all drivers? This includes GPU, MB / Chipset drivers from the manufacturers website? (Don't use third party driver installers).

Do you have latest BIOS installed?

Overall asides from the above:
I already lowered the cpu to 3.4 GHz and put the Ram to 3.2 GHz (CL 16).
I would reset BIOS completely to default first. Then retry. Just being everything back to stock setting.

When I say it crashes I mean PC turns off (Blackscreen) and tries to reboot afterwards. It usually needs several tries (further crashes) until it gets back to a stable desktop.
It looks like it could be either GPU or PSU.

My current thoughts are: Is the PSU to weak ? But 650 W should be okay right ?
Presumably these are new parts? 650W is plenty.
The Seasonic Focus Plus Gold are excellent PSUs too.

Or is the Graphics card broken ? But how could i then just write this Post here on the Computer ? System is perfectly stable unless is stress GPU
I'd rest my bets on GPU initially. So I'd probably return that. Following that, PSU. Better if you can simply try another similar GPU in your system and see if the issue persists.
 
Aug 25, 2019
15
2
15
0
Hi Guys,

i just built a new system:

AMD Ryzen 7 3700X
MSI MPG X570 Gaming Plus
16GB G.Skill Trident Z DDR4-3600 DIMM CL15 Dual Kit
8GB Sapphire Radeon RX Vega 56 Pulse Aktiv PCIe 3.0 x16
Seasonic 650 FX Focus+ Gold

Windows 10 is fully upgraded.

But when I try to play a Game the Computer shuts off after like a minute.

I already lowered the cpu to 3.4 GHz and put the Ram to 3.2 GHz (CL 16).
But this didnt solve the issue.
To nail the problem down I used CPU-Z to stress CPU. -> Nothing happens ... runs stable
Then I used memTest (6 of them 2.5Gb each) to stress RAM. -> Nothing happens ... runs stable
Then I used FurMark 3D to stress GPU. -> System crashes after like 20-40 seconds.

When I say it crashes I mean PC turns off (Blackscreen) and tries to reboot afterwards. It usually needs several tries (further crashes) until it gets back to a stable desktop.

When the crashes occured the GPU had 60° C or 140° F according to WattMan.

I am really struggeling to find out whats the issue here guys.

My current thoughts are: Is the PSU to weak ? But 650 W should be okay right ?

Or is the Graphics card broken ? But how could i then just write this Post here on the Computer ? System is perfectly stable unless is stress GPU

Any advice would be really helpful.

Thanks in advance :)

Mario
 

PC Tailor

Distinguished
Herald
Welcome to the forums my friend!

As a first port of call, have you updated all drivers? This includes GPU, MB / Chipset drivers from the manufacturers website? (Don't use third party driver installers).

Do you have latest BIOS installed?

Overall asides from the above:
I already lowered the cpu to 3.4 GHz and put the Ram to 3.2 GHz (CL 16).
I would reset BIOS completely to default first. Then retry. Just being everything back to stock setting.

When I say it crashes I mean PC turns off (Blackscreen) and tries to reboot afterwards. It usually needs several tries (further crashes) until it gets back to a stable desktop.
It looks like it could be either GPU or PSU.

My current thoughts are: Is the PSU to weak ? But 650 W should be okay right ?
Presumably these are new parts? 650W is plenty.
The Seasonic Focus Plus Gold are excellent PSUs too.

Or is the Graphics card broken ? But how could i then just write this Post here on the Computer ? System is perfectly stable unless is stress GPU
I'd rest my bets on GPU initially. So I'd probably return that. Following that, PSU. Better if you can simply try another similar GPU in your system and see if the issue persists.
 
Aug 25, 2019
15
2
15
0
Thanks for the help so far. Drivers are up to date. BIOS was not. After the update everything seems more stable. I will let you know when I had some more time to test the system.

Thank you
 
Reactions: PC Tailor
From your description this indeed sounds like a PSU issue to me, though that Seasonic 650 FX Focus+ Gold PSU is a very good power supply. Even high quality units are prone to failure. but the chances are less.

Have you overclocked your GPU ? Have you tried undervolting that VEGA GPU ? Just make all the PCI-E power cables are properly connected to the GPU, and also try to reseat the Video card on the primary PCI-Express x16 slot.
 

ASK THE COMMUNITY