Last night I experienced a crash that spiked the fans and locked the system on a black screen. Now in game I get artifacts and flickering across multiple monitors and then it crashes again. In diagnosis I found the said high junction temp under load. The main temp and memory temp readings barely got above 60°C. I know junction temps aren’t going to be the same but that’s a very large delta no? Am I correct in assuming this is some sort of hardware failure? It’s a Gigabyte Radeon RX 6750 XT for reference
If your card is still covered under warranty try starting an RMA. If you are more tech savvy and don’t have warranty you can try to:
It’s supposed to be a “feature” that they ramp all the way up to 110 C at the junction, nothing to be alarmed about…
I had the same issue with my 6800 XT, but can keep the thermals down to 70° with undervolting.
Was that gigabyte’s answer to your rma request?
If that’s the case you should follow up while playing dumb like “it got worse. every game crashes after 2 minutes”. No need for specifics, just exaggerate the symptom a bit.
High junction temp indicates wrong cooler contact or paste dry out. You can’t fix those without damaging your “warrany void” sticker.
They might try to argue against you on it, but it’s at least worth noting that “warranty void if broken” stickers haven’t been legal/enforceable since 1975. The FTC started going after companies a bit more about it starting back in 2018.
Took em a while. Lol