RTX 2080Ti less prone (maybe) to crashing on 80% power target

[crispy.ltt]

Hi, I've been experiencing some issues with my 2080 Ti and I've tested just about everything at this point. TLDR:

1. Completely different system
2. Different PSU
3. Different PSU cables
4. Single daisy chained 2x8pin vs. two separate 8pin
5. Reducing power target to 80% (thought this fixed it, but now i am getting crashes here as well)
6. Improved cooling, max 65C on core 82C on hotspot (using testbench and external fans blowing on card)
(even at 130% power target and +150Mhz core +1000Mhz memory)

The crashing is very hard to diagnose. They look like VRAM failure but occur very strangely. I have ran 15 hour 3Dmark benchmark loops with no crashes, but in various games I can get crashing within 30s or after 3 hours. It has however crashed in benchmarks as well. (I think) the card crashes less on a lower power target, but this seems to also have caught up with me. I now see frequent crashes on 80% power target as well.

HWinfo reports a couple HWEA PCIe failures sometimes.

Not sure if I can save this card or not. Last resort is to mount a waterblock to it, but I doubt heat is the issue given how cool it runs even at higher OC and power targets.

[RevGAM]

1 hour ago, crispy.ltt said:

Hi, I've been experiencing some issues with my 2080 Ti and I've tested just about everything at this point. TLDR:

1. Completely different system
2. Different PSU
3. Different PSU cables
4. Single daisy chained 2x8pin vs. two separate 8pin
5. Reducing power target to 80% (thought this fixed it, but now i am getting crashes here as well)
6. Improved cooling, max 65C on core 82C on hotspot (using testbench and external fans blowing on card)
(even at 130% power target and +150Mhz core +1000Mhz memory)

The crashing is very hard to diagnose. They look like VRAM failure but occur very strangely. I have ran 15 hour 3Dmark benchmark loops with no crashes, but in various games I can get crashing within 30s or after 3 hours. It has however crashed in benchmarks as well. (I think) the card crashes less on a lower power target, but this seems to also have caught up with me. I now see frequent crashes on 80% power target as well.

HWinfo reports a couple HWEA PCIe failures sometimes.

Not sure if I can save this card or not. Last resort is to mount a waterblock to it, but I doubt heat is the issue given how cool it runs even at higher OC and power targets.

Have you tried repasting it and/or putting in pads instead?

[crispy.ltt]

15 hours ago, RevGAM said:

Have you tried repasting it and/or putting in pads instead?

I have repasted it, but I havn't changed the thermal pads. You think it might be overheating?

[D2ultima]

19 hours ago, crispy.ltt said:

Hi, I've been experiencing some issues with my 2080 Ti and I've tested just about everything at this point. TLDR:

1. Completely different system
2. Different PSU
3. Different PSU cables
4. Single daisy chained 2x8pin vs. two separate 8pin
5. Reducing power target to 80% (thought this fixed it, but now i am getting crashes here as well)
6. Improved cooling, max 65C on core 82C on hotspot (using testbench and external fans blowing on card)
(even at 130% power target and +150Mhz core +1000Mhz memory)

The crashing is very hard to diagnose. They look like VRAM failure but occur very strangely. I have ran 15 hour 3Dmark benchmark loops with no crashes, but in various games I can get crashing within 30s or after 3 hours. It has however crashed in benchmarks as well. (I think) the card crashes less on a lower power target, but this seems to also have caught up with me. I now see frequent crashes on 80% power target as well.

HWinfo reports a couple HWEA PCIe failures sometimes.

Not sure if I can save this card or not. Last resort is to mount a waterblock to it, but I doubt heat is the issue given how cool it runs even at higher OC and power targets.

Have you tried putting it in a different slot on your motherboard if it exists?

[RevGAM]

6 hours ago, crispy.ltt said:

I have repasted it, but I havn't changed the thermal pads. You think it might be overheating?

Given its age, maybe. 

[crispy.ltt]

On 10/11/2023 at 10:28 AM, D2ultima said:

Have you tried putting it in a different slot on your motherboard if it exists?

Yes, i tested it on different slots with my older Z97 ATX as well as my new X570 ITX motherboard.

 

On 10/11/2023 at 2:24 PM, RevGAM said:

Given its age, maybe. 

Ok, i'll order some new pads and see if that makes a difference.

[D2ultima]

4 hours ago, crispy.ltt said:

Yes, i tested it on different slots with my older Z97 ATX as well as my new X570 ITX motherboard.

 

Ok, i'll order some new pads and see if that makes a difference.

Your card is probably dead bro

 

At this point I'd take it to a shop that can electrically test it for faults instead of trying to blindfire it.

[AI_Must_Di3]

1 hour ago, D2ultima said:

Your card is probably dead bro

 

At this point I'd take it to a shop that can electrically test it for faults instead of trying to blindfire it.

I agree. Make sure the new pads you get are the right size cause its crucial. Too fat and itll make the HSF sit too high and other stuff may not get cooled. Too thin and youll miss the pads completely.