Weird 2080ti behavior

Olivia Moore · August 1, 2019

Hey guys. So I installed after burner to adjust the fan curve on my 2080ti get so it would be so hot while running games. I was satisfied that the curve brought my temps down from 82c to around 70-73c.

I have not applied any over clocking to the card in after burner. This morning I started my rig up and started fort nite for it to crash, then I started the evil within 2 and that crashed. For a second before it crashed in the EW2 I saw some “Christmas Lights” but in fortnite before it would crash I didn’t see any.

I had to do a hard reset after the ew2 crashed. After that, it seems to be working fine. Fortnite and ew2 plays with no problem. I even ran heaven benchmark for a little bit after the crash and it looks fine too.

So this a fluke? Driver malfunction maybe? Should I be thinking about RMA?

AGAIN I have NOT applied any overclocks in after burner, just a custom fan curve.

Olivia Moore · August 1, 2019

4 minutes ago, Genwyn said:

furmark it for a bit, see what happens

if its artifacting at all it could be toast

try the regular lineup of driver update, maybe try an older driver version as well, return it to all stock settings (even if that means hotter than before), and since its happening as youre in a game it may be power related, make sure your PSU is up to the task, try lowering the clocks to reduce power consumption and see if the issue persists

I should have stated before I have a Corsair rm850x that’s about 8 months old.

Could have setting that fan curve messed with power consumption- voltages?

Olivia Moore · August 1, 2019

47 minutes ago, Genwyn said:

nah, not drastically, the fans dont use enough power or run at a high enough voltage to cause issues on the board

theres a million little tiny things that it could be, the artifacting is the most concerning bit because usually a GPU driver crash wouldnt result in any artifacting, artifacts are almost always a hardware issue

again, furmark it, try different driver versions, play around with the clock speeds a bit just to see if maybe its something small like the memory not liking whatever its stock clock speed is (in that case, RMA it)

if it doesnt stop, RMA

I will try furmark later when I have more time to stay around it while it is running.

Before I had to leave I played a match in fortnite, I Have most of my settings set to high. And I did not see any artifacts or crashes. I know that may not mean anything though and the card could still be going bad. But I guess I am hopeful that I didnt see anything wrong.

Olivia Moore · August 2, 2019

@Genwyn Update: So I never did run furmark, but I was able to play games yesterday for a few hours with no crashes or artifacts. It acted normal. All memory and boost clocks stable and temp was good. It also does fine in heaven benchmark.

Princess Luna · August 2, 2019

11 minutes ago, Olivia Moore said:

@Genwyn Update: So I never did run furmark, but I was able to play games yesterday for a few hours with no crashes or artifacts. It acted normal. All memory and boost clocks stable and temp was good. It also does fine in heaven benchmark.

FurMark is uselss any ways, its workload was so unrealistic and demanding that it used to literally kill GPUs back on AMD Hawaii and nVidia Maxwell days... the problem was such that both nVidia and AMD had to include micro-code on Pascal and Polaris/Vega that when detected FurMark the card throttles as heavy as possible to protect itself.

So this application really doesn't serve any purpose from how far it differs from gaming loads and how the GPU is set to purposely throttle on it, trying to check an OC stability on it is meaningless, Heaven or something more recent like Superposition is far more adequate.