Hello, I'm relatively new to the world of PC gaming (2 years), and hoping to get some advice with respect to temperature issues I've been having with my MSI Ventus RTX 3080. (Yes, I am also a luckbox, got this through Newegg in October 2020 before the GPU availability got really bad).
Basically, I am having an issue where my GPU temperatures, when gaming, are slowly getting worse after about a month from replacing the thermal paste, at least according to what hwinfo64 and GPU-z are reporting.
As with other people who have this and other similarly-constructed RTX 3080s, I replaced the stock thermal pads for the VRAM modules to improve VRAM temperatures, and since then I have zero issues with the VRAM temperature (usually around 75-78 degrees Celsius after prolonged periods of gaming).
When I replaced the thermal pads, I repasted the GPU using Noctua NT-H1, and at first my GPU and hotspot temperatures were amazing (GPU maximum was 68 degrees, hotspot around 80). However, after about 2 months, I noticed my fans were a lot louder than before, and sure enough, the GPU started hitting around 78 while gaming, the hotspot temperature was consistently around 96-97, and the fans are often running at or near 100% speed.
I wasn’t sure if it was maybe an issue with the thermal paste, as I had seen some people report in various forums that thicker thermal paste often works better for GPUs. So I replaced the thermal paste, this time using Artic MX-4. This time, once again, my temperatures were much better during gaming, 68 degrees max GPU temperature, 78 degrees hotspot temperature, and fans weren’t going any higher than 80%. But, once again, after about a month, I noticed the temperatures and fan noise creeping up again.
Currently, while gaming the GPU temperature sits steadily around 74 degrees, and the hotspot is around 94, and occasionally the temperatures spike to 78 and 98 respectively. When idle, GPU temperature is usually at 35 and hotspot at 45, and during light use (such as zoom meetings) the temperatures are around 45 and 55, so there is a pretty consistent 10 degree difference, except when gaming. Based on the information provided through hwinfo64 and gpu-z, the card isn't thermally throttling.
So, my questions are:
1. Is it normal to see behavior like this after replacing thermal paste in a GPU, or is the thermal paste degrading prematurely for some reason?
2. Are these temperatures even something I should worry about for an RTX 3080?
3. If they are, do people have recommendations for a different thermal paste for this application?
Thanks in advance, and again, I just want to be clear: in both cases, the GPU temperatures were initially totally fine for at least a month after repasting, so I don’t think this is an issue of poor contact with the heatsink, the thermal pad replacement, a missing thermal pad, using replacement pads that are too thin or thick etc. I only replaced the pads for the VRAM modules, and even with higher GPU and hotspot temps, the VRAM temperatures reported in hwinfo64 are totally fine.
Setup (in case any of this is relevant:
RTX 3080
Core I9-10850k
Memory: 32 GB Ram 3200 MHz, CL 16
Gaming resolution: 3440 x 1440p