Question Crash : Randomly getting blackscreen RTX 3070 ti

[Arislas]

So, as the title said, my problem is that at some random moments, my GPU will kind of crash.

My specs :

• CPU: Intel i5-12600 3,3 GhZ( 4,8 Ghz Turbo Boost) socket 1700
• RAM: Corsair Vengeance LPX 32GB (2x16GB) DDR4 3200MHz C16 - Black
• GPU: GIGABYTE GEFORCE RTX 3070 TI GAMING OC 8GD
• PSU: NZXT C SERIES C750
• Mobo: Gigabyte B660 Gaming X DDR4
• Case: Be quiet Pure Base 500DX RGB

For the story :
This is a computer I assembled myself few months ago. It worked really well, until I move it to my friend place for a LAN party, there it started to crashed randomly while playing and then even not playing ( just after booting on windows, or even not booting at all)

So far, the problem I saw are :

• Sometimes, my GPU crash ( while playing or even just on windows desk nothing running : GPU led color freezing, no more output on my screen. The computer will not reboot anytime but sometimes yes. The crash can occur after 30 seconds or 1 day.
• After rebooting after this kind of crash , I won't get any output from my GPU as my Motherboard will show me there is a problem with my GPU ( VGA Led on )
• The graphic card get hot and will start full speed fan mode after a few minutes after a crash.
• I recently discover that doing any movement to my GPU will cause a partial crash ( monitor going off, sound start to freeze, then after a minutes it will go back on )

As all my component are still under warranty, I need to know where is the problem exactly. I don't want to sent back my GPU to find out with the new one that the problem is from my PSU or Motherboard. So I tried a few things :

• Run my computer with an other GPU ( 1080 ti ) : all is clean even with a full stress test ( gently shake the GPU )
• Run my older computer with my new GPU, but I can't get any output to my monitor ( may be cause by my old PSU ( 8 years old ) )
• Windows 11 update and Nvidia driver update

I really think that the problem is my GPU but I really want some opinion from other people here. Ask your question and give me some things to try out !

 

[KyaraM]

Sounds like your GPU's connector to the mainboard might have gotten damaged during transport? Since you tested with an old card and presumably reinserting the 3070Ti into the PCIE slot didn't fix the problem, I would still check the CPU cooler that it sits firmly ober the CPU, and take out the RAM sticks and put them back in, just in case. Also, check all cabling both at the PSU end and at the components including the GPU. Make sure everything sits firmly. If nothing helps, you probably have to send it in for a check...

 

[Arislas]

Yeah, I think it could be the connector too ! Just not sure it's the GPU's one, could be the MOBO's one ( but as the other card work well it shouldn't this one )

Yes, I check a few time if my GPU was well inserted, and the cable or good too, I even switch those for the GPU with new ones.

I don't understand why I should check the CPU cooler ( CPU fan in my case ). The CPU fan is firmly screwed to the motherboard as I used an ARCTIC Freezer i35. So I don't see what I can really check.

For the RAM, I already try to switch those components at the beginning before I discover that it was a GPU problem. As my crash are terribly random, I can really check them properly. Except by shaking my GPU, which will cause the crash I think. Good or bad RAM.

 

[Arislas]

I just find out with the reliability monitor that I got some Bluscreen behind my blackscreen. There is the information :

Signature du problème
Nom d’événement du problème : BlueScreen
Code: 116
Paramètre 1: ffff968b357a1370
Paramètre 2: fffff80430e91218
Paramètre 3: ffffffffc000009a
Paramètre 4: 4
Version du système d’exploitation: 10_0_22000
Service Pack: 0_0
Produit: 768_1
Version du système: 10.0.22000.2.0.0.768.101
Identificateur de paramètres régionaux: 1036

Informations complémentaires sur le problème
ID du récipient : 0x116_TdrBCR:4:C000009A_Tdr:A_IMAGE_nvlddmkm.sys_Ampere

 

[KyaraM]

That last line sounds like the Invidia driver. Did you try to uninstall it in save mode using DDU, then reinstall?

The problem appeared here

https://www.nvidia.com/en-us/geforce/forums/geforce-graphics-cards/5/489240/videotdrfailure-116-attempt-to-reset-the-display-d/

and here the culprit was iCUE

**SOLVED** - VIDEO_TDR_FAILURE error 116 - Please help I'm at wits end

Hopefully this is somewhat the correct place for this but this is my first ever post on a forum so please forgive me if I have made a mistake. I'm at wits end with an issue that's been giving me grief for months now. I have been plagued with BSOD (doesn't actually bluescreen but I can see in...

forums.tomshardware.com

 

[Arislas]

Well, I'm actually saving all the file I need to keep. Then I will try to uninstall the Nvidia Driver and also repairing my windows ( my ethernet port isn't working anymore and it doesn't appear in the device manager)

 

[Arislas]

I'm currently reinstalling the driver. And it makes me think that after the beginning of the crash, I updated my Nvidia drivers. So the crash began before that :/ I even updated Windows 11

 

[Arislas]

Well, It is not the driver as I got a crash while I have no Nvidia driver at all.

And I also got a new problem : my ethernet port isn't working anymore. Device manager say that the device is not connected at all.

 

[Arislas]

So... Some news :

I sent my motherboard and graphic card to repair and I got it back yesterday. My graphic card is fine they said, and they changed my motherboard. But, after I built my computer again, I still got the black screen, just before the boot. No output on the screen and the motherboard red light is on for VGA.

So, I brought my computer to my friend's house, and we tried some things :

• Put my graphic card on his computer: it worked well.
• Put his graphic card on my computer ( 3080 ): it worked well.

With that, I'm totally lost. The only thing left that, for me, can cause any issue, is my PSU. But why does it work well with my friend's GPU and not mine ? His GPU consume even more power than mine.

 

[Alan Alan]

Video cards are heavy, pull the card and look, did you crack the board, does the connector flex. Try a different pcie slot if you broke one.

 

[KyaraM]

Video cards are heavy, pull the card and look, did you crack the board, does the connector flex. Try a different pcie slot if you broke one.

I mean, the board was replaced already and the issue persists. Seems unlikely that the slot broke on two boards back to back, and basically immediately...

 

[Alan Alan]

I mean, the board was replaced already and the issue persists. Seems unlikely that the slot broke on two boards back to back, and basically immediately...

I mean, the board was replaced already and the issue persists. Seems unlikely that the slot broke on two boards back to back, and basically immediately...

I agree, perhaps the Gpu has a cracked board. On the other hand send it back while it's under warranty. It could have something wrong with it. Sometimes things have loose connections and we never know until the get moved around. Most definately send it back, they will repair or replace it at their option. Sometimes solder doesn't take well on hot running parts and the legs pop loose. If it's a common problem they will have a fix for it. Good luck.