A sizzling potato: It is laborious to consider that Nvidia’s RTX 4090 turned one 12 months outdated at the moment. Perhaps that is as a result of each few months, Meltgate rears its ugly head. Sure. One other consumer has reported that his 4090 dedicated suicide after working simply tremendous for a 12 months.
Simply whenever you thought it was secure to make use of your 12VHPWR cable together with your RTX 4090 once more, one other incident of GPU meltdown pops up within the boards. Redditor Byogore stories that he purchased an Asus 4090 in Germany a 12 months in the past due to US provide points. The cardboard labored completely tremendous till it self-immolated two days in the past. This incident is uncommon as a result of sometimes failures have occurred a lot sooner.
Different Redditors shortly questioned whether or not Byogore had the connectors seated securely since “consumer error” was one in every of Nvidia’s excuses when the difficulty arose shortly after the RTX4090 launch. Byogore defended his potential to correctly connect a pc element.
“Clearly, I did…” he identified. “If this was a seating subject, it will’ve died ages in the past. I’ve used it loads.”
He additionally pointed to a video from a California restore store, NorthridgeFix, that stated it has seen many 4090s spontaneously combusting (under). Primarily based on inside assessments, NorthridgeFix is adamant that the issue just isn’t the cables, connectors, bending, PSUs, or consumer error.
“The truth that the 90-degree cable mod adapter [to prevent bending] was plugged in totally to the connector and the connector nonetheless melted, then we all know that we have now an issue with the cardboard. Now we have an issue with the cardboard. We don’t have an issue with the cable. It is not consumer error, it isn’t a cable not plugged in correctly downside, it isn’t a cable mod downside, but it surely’s an issue with the design and the engineering of the cardboard.”
NorthridgeFix additionally factors out that even when the issue was brought on by consumer error, it is nonetheless Nvidia’s fault. As an example, if a consumer is plugging in a cable as they do with every other, previous or current, and it leaves a 1mm hole that causes failure, that isn’t the consumer’s fault. When there’s a tolerance subject like that, it’s the producer’s duty to repair it or design a mechanism to forestall it, not blame the consumer for not plugging it in appropriately and doing nothing extra about it.
Byogore says that the meltdown occurred whereas he was taking part in Battlefield 2042. The display turned black, however the audio continued. Then, the pc rebooted itself. Because it began up, he may scent burnt plastic. The cardboard nonetheless labored, however solely briefly earlier than crashing and burning once more. Byogore talked about that he has a 1,000W Corsair PSU and that the 4090 was undervolted on the time of the catastrophic failure. Nonetheless, he did not word whether or not his PSU was ATX 3.0. The issue solely appears to happen on older ATX 2.0 energy provides.
Fortuitously, Asus has been very empathetic and beneficiant to Byogore’s plight. When he contacted customer support, they provided to improve him with a brand new Strix card (from a TUF mannequin) or give him a full refund. Presumably, the cardboard was nonetheless below guarantee because it was two days shy of its first anniversary.
Nvidia has been reluctant to just accept any blame for the 4090’s hot-button woes. It initially blamed customers for not securing the cable tightly to the socket. Later, it stated that “poorly designed” 12VHPWR adapters have been responsible. The final time the difficulty popped up sufficient to make information was final Could. The continued downside has even sucked Nvidia right into a class-action lawsuit.