A user discovers a missing thermal pad in the VRAMs of his Asus RTX 3080 Ti TUF OC

Reddit user kamaloo92 discovered that a thermal pad was missing from the GDDR6X VRAMs of his Asus RTX 3080 Ti TUF OC after noticing that its graphics suffered from severe overheating that led to a noticeable drop in FPS in games. The user first performed an exhaustive complete software review, and after seeing that everything was in order, he decided to dismantle the graphics card to check the hardware, finding this surprise.

My RTX 3080 had you a very high VRAM temperature, almost always 110 ° C. So I finally opened it to find that a thermal pad was missing on four VRAM chips.

There are some marks on the heat sink, as if the pad has been in place for a few seconds, but the VRAM chips are clean. So I guess the thermal pad somehow fell off before I put the heatsink on the board.

It worked normally, without any kind of reboots, it only suffered constant thermal throttling.

Other Reddit users asked him which heating pads he used as a replacement. It is common practice for more experienced users to replace heated pads for better performance., although in this case it was not for improvement but for necessity. In this regard, kamaloo92 explained:

I added the missing pad and replaced the others, just to be safe. I also added additional pads as recommended for the 3080 TUF (not Ti) and the temperatures dropped to a maximum of 80 ° C. So I hope everything works fine.

Just in case, I have used Gelid GP-Extreme. 1.5mm between the heatsink and VRAM, 1.5mm at the end of the video connectors, 3mm between the heatsink and the main heatsink, 3mm at the end of the power connector. I haven’t removed the back plate so I don’t know what’s there. On the GPU I have used Kryonaut.

The RTX 3080 Founders also have problems with the thermal pads, although in this case, due to poor placement:

In the same Reddit r / nvidia forum, 7 days ago the user obamaprism3 discovered a defect with the thermal pad that should cover the VRAM of your RTX 3080 Founders Edition, since it was running and was resting on the area that should make contact with the GPU, causing it to lift up and not make proper contact with the GPU, and therefore the graphics to overheat.

These are some of the many cases that have appeared in recent months, where manufacturing defects or objects are found in some pieces of hardware. It is speculated that these problems may be the product of the pandemic: over-demanded employees on the assembly lines, perhaps a workforce without the necessary experience hired to cover the lack of employees. Beyond the cause of these problems, it is important to be aware of the temperatures of our hardware for possible similar problems.

Did you ever have factory issues like these with your GPUs?

Source: WCCFTech