[SOLVED] Graphics Card Disconnects for half a second and Reconnects.

octaゼ永あ

Prominent
Dec 26, 2020
7
1
515
Hi!
I'll start showing my PC hardware, in case you guys can see the problem there.

-CPU: Ryzen 5 1600AF

-GPU: GTX 970 MSI Gaming (OVERCLOCKED) 160+ Core Clock, 500+ Memory Clock


-Motherboard: Asrock B450M Hdv

-Ram: (2x8) 16GB Oloy Warhawk 3200mhz

-PSU: Thermaltake Smart 500W 80Plus

-SSD: 480GB KINGSTON A400

-HDD 1: WD BLUE 1TB

-HDD 2: WD BLUE 500GB

My problem is that whenever I play a game, after half an hour or more, it crashes (but I still hear the sound of the game) showing a black screen or a solid color screen following the palette color of the game that i'm playing (weird, by example, orange color when I'm playing csgo and dark blue or grey when I'm playing Dark Souls III) I didn't knew why that happened, and I investigated and found out that is because the GPU disconnects, also I found out that the GPU disconnects because one game that I played showed an error message saying "Unreal Engine is exiting due to D3D device being lost" is not an Unreal Engine problem for me, happens at almost every game, doesn't matter if it's demanding or not. That problem happens too when I'm on youtube but doesn't crash the browser or anything.

-My graphics card is not overheating, the max temps that I got are 76c°
Here I show my temps while I'm playing Sekiro, https://ibb.co/KX9RqF5

-It's not happening because of the overclocking, I oc'ed 5 or 6 months ago and had no problem.

-I already reinstalled the graphics drivers, I'm using windows 11 right now and I made a clean install of both

Also, I took an screenshot when the crash happened, It was showing a green screen but the screenshot shows a black screen.
Here it is https://ibb.co/zf4jJ7V

I did found a video on youtube of the problem but seems like it was a loose cable on a modular PSU, mine is not modular but I already tried disconnecting and reconnecting the 8 pin and 6 pin cables from the GPU but looks like the problem still persist.

I really hope that is a loose cable or something, I can't afford any card right now.
Sorry if my English is not that good, I am not a native English speaker.

I could really use your help right now! Thanks in advance
 
Solution
First you should revert OC and run just with the default clocks. Just because an OC run for an X amount of time does not mean that it will continue to run stable , not even for the next hour. Overclocking puts extra stress in the GPU which does affect the silicone and shorten it's lifespan and clocks achieved.

Then you got that terrible PSU. It's 420w in reality and that under ideal conditions. It's garbage quality and you pushed it even further by OCing the GPU. Using it was bad enough...

The PSU might have damaged the GPU. You need to replace it asap and then continue any troubleshooting.

You can also try DDU in safe mode but I wouldn't have my hopes high...
First you should revert OC and run just with the default clocks. Just because an OC run for an X amount of time does not mean that it will continue to run stable , not even for the next hour. Overclocking puts extra stress in the GPU which does affect the silicone and shorten it's lifespan and clocks achieved.

Then you got that terrible PSU. It's 420w in reality and that under ideal conditions. It's garbage quality and you pushed it even further by OCing the GPU. Using it was bad enough...

The PSU might have damaged the GPU. You need to replace it asap and then continue any troubleshooting.

You can also try DDU in safe mode but I wouldn't have my hopes high.
 
Solution

octaゼ永あ

Prominent
Dec 26, 2020
7
1
515
First you should revert OC and run just with the default clocks. Just because an OC run for an X amount of time does not mean that it will continue to run stable , not even for the next hour. Overclocking puts extra stress in the GPU which does affect the silicone and shorten it's lifespan and clocks achieved.

Then you got that terrible PSU. It's 420w in reality and that under ideal conditions. It's garbage quality and you pushed it even further by OCing the GPU. Using it was bad enough...

The PSU might have damaged the GPU. You need to replace it asap and then continue any troubleshooting.

You can also try DDU in safe mode but I wouldn't have my hopes high.
Hi there, well, I'll try using without OC to see if it works.

Sorry but I can't afford a new PSU right now, I'll use it without Oc to try to drain less power.

I already used NVcleaninstall but the problems were still the same.
Thanks a lot!
 

tecmo34

Administrator
Moderator
In addition, you may want to switch out your GPU cable with a new one, if possible. It could be the cable from the monitor to the GPU is going bad, so it goes in and out. I had this happen when I tried to run a DP extension, as my original DP cable wasn't long enough. Once I bought the correct length, I haven't had issues since.
 

octaゼ永あ

Prominent
Dec 26, 2020
7
1
515
I think my problem might be solved, my PSU couldn't deliver all the power to the graphics card, also I got too much things connected to my motherboard (I don't know if that affects the power delivery in someway ) but I didn't had a crash since yesterday, it used to be often. I'll try to buy another PSU, they're a lot less expensive than a decent GPU.
Thanks! to all of you!
 
  • Like
Reactions: dotas1