Posts: 86 +40
Over the past few weeks, reports have popped up of dying GeForce RTX cards without any particular reason whatsoever. The problem revolved mostly around Founders Edition RTX 2080 Ti cards, although some AIB partner cards were also met with similar issues. The issue manifested with graphics artifacts on-screen just before freezing up the user's PC and requiring a cold boot to get it back up and running.
The artifacts look like someone's overlaying Space Invaders on your game, only the end result is a BSOD. Image credit goes to GeForce forum member SawFEAR.
Several media outlets ran their own investigations into the matter (including Gamers Nexus who has asked owners of dead RTX 2080 Tis to ship them their way), with many users pointing an accusing finger at Micron's GDDR6 memory modules as the culprit. This was further (accidentally) fueled by Nvidia who, in the meantime, started shipping a new batch of RTX 2080 Ti using Samsung-sourced GDDR6. This was written off later as 'business as usual', as the supply of GDDR6 chips from Micron was now non-exclusive, allowing Samsung (and maybe SK Hynix, down the line) to join the party. As time went on, it became clearer that the issue related almost exclusively to Founders Edition cards, which -- adding insult to injury -- are the ones that carry a $200 premium over the reference pricing.
With nothing happening on the Nvidia side and users screaming on Reddit and GeForce Forums until their faces turned red, not to mention the humiliation of spending $1,200 on a borked gaming GPU, it's finally addressed the issue... sort of.
In a rather short and low-key forum post, Nvidia has addressed the issue and posted the following:
"Limited test escapes from early boards caused the issues some customers have experienced with RTX 2080 Ti Founders Edition. We stand ready to help any customers who are experiencing problems. Please visit www.nvidia.com/support to chat live with the NVIDIA tech support team (or to send us an email) and we’ll take care of it."
While fairly succinct we can read a bit into the statement: First of all, it's a sure bet Nvidia knows, by now, what is going on and new batches will no longer have the issue, otherwise it would just be inviting itself a world of hurt.
According to users on Reddit, their RMAs are arriving with new RTX 2080 Ti cards and they work just fine. Secondly, the issue seems restricted to the RTX 2080 Ti, which makes it unique to some feature that is not shared with the 2080. Since the RTX 2080 Ti and non-Ti cards share the same memory controller and chips, it's highly unlikely this was caused by a hardware defect in those components (I have a feeling these are power issues). Third, "test escapes" are a way of saying that the manufacturing missed some point of the quality assurance process and a batch of cards went untested (most likely due to machine error, as human error would end up affecting a smaller-than-reported number of cards).
We've reached out to Nvidia for comment on the situation, but so far no luck, so check in later for possible updates. If you're one of the unlucky souls who've had their RTX 2080 Ti crash on them, only to find yourself on the phone with Nvidia tech support, let us know your experience in the comments. Nvidia seems to be resolving the issue for customers without much fanfare, and these things do happen when something is rushed out of the door.