flyinion

Sajin
I just max it out for extra stability. If you don’t black screen you’ll know you weren’t stable.

What setting are you putting the voltage control to in order to get 1.1mv?  It's showing as a % so I'm thoroughly confused and don't want to overvolt the card.  For the dropdown did you leave it on "Standard MSI" or put it on 3rd party or reference?

flyinion

Sajin
Sounds like the extra voltage helped you become stable. Nvidia does blacklist specific apps to keep them from using max voltage like furmark, so it looks like occt is one of those apps.

Try running cp2077 with the voltage slider maxed out, and you’ll definitely see 1.1mv being used.

Hmm not sure on that since I put it back to default after.  I think it was a background thing/etc causing a conflict.  Unfortunately I don't own CP2077 so can't really test with that.

flyinion

Sajin

Any demanding game should do the trick then.

Have you tried running OCCT basic GPU test at the......I think....default load of 100% with your voltage settings or just stock since your repaste?  I'm curious to find out what your hotspot delta is on that vs the main GPU temp.  I had OCCT set to stop any test if a sensor hit 80C and about 7.5 minutes into a 100% load GPU test it shut down due to 80C on the hotspot with 61 or 62 on the core.  Water was up to 37C but I didn't have my fans fully ramped.  Pretty sure it's passed this test before though so I'm wondering if TG dries out, or my mounting screws are no longer as tight as they should be, the 4 around the core.  I don't remember what cooling you're on though, in my case I'd have to try and get an EKWB backplate off while it's in the rig on a vertical GPU mount.  Temps are usually within expected deltas in actual content though.

kraade
600W to hit 80  and now its hitting it at 480, I think maybe the vapor chamber might be compromised. Unless someone who blocked a card sells you a stock cooler , RMA might be the only choice 

flyinion

Sajin

rjohnson11
Good Luck and hope this gets solved quickly

It’s resolved for now. Card is working fine at the moment. If it gets hot again I’ll try some thermal grizzly kryonaut, or maybe even some kingpin cooling kpx before rma’ing.

So I"m not sure about the paste you did have on there, but I was debating giving KPX a try and was looking up comparisons to Kryonaut which I've been using.  Found some info that apparently Debauer (spelling?) has said Kryonaut on GPU's is expected to be replaced after a year.  If true that might explain my hotspot temp also seeming to be a bit higher under OCCT type loads than I remember.

cesarx

Sajin

flyinion

Sajin

rjohnson11
Good Luck and hope this gets solved quickly

It’s resolved for now. Card is working fine at the moment. If it gets hot again I’ll try some thermal grizzly kryonaut, or maybe even some kingpin cooling kpx before rma’ing.

So I"m not sure about the paste you did have on there, but I was debating giving KPX a try and was looking up comparisons to Kryonaut which I've been using.  Found some info that apparently Debauer (spelling?) has said Kryonaut on GPU's is expected to be replaced after a year.  If true that might explain my hotspot temp also seeming to be a bit higher under OCCT type loads than I remember.

Before I repasted it was the original paste. Not sure what the original paste was, but it was really good until it got burnt off. When I repasted there was hardly any of it left on the die.

try honeywell ptm7950 its pretty good

safan80

Sajin
I would never put liquid metal on my hardware. 😊

Sony uses it on the PS5... That is kind of scary.