[SOLVED] MSI GeForce RTX 2070 SUPER VENTUS GP OC - Throttling at 70c

Oct 11, 2020
2
0
10
I have a MSI GeForce RTX 2070 SUPER VENTUS GP OC that is throttling at 70c. This has cut my performance down 20-30%. Below is a background of how this came about.

The cards runs hot as many know. In attempt to cool things down I replaced the thermal paste and pads. After this the card hit the limit set of 83c very quickly in my initial tests. I tried re-apply thermal paste and pads a couple times with no success. Rather than run the computer and risk damaging things I shut it down and decided to try different cooling options. After reading an article here I ordered a Innovation Cooling Graphite Thermal Pad for the GPU and Thermal Grizzly Minus Pad 8 Thermal Pads to replace the pads on the other chips.

After this the GPU started throttling at 70c. Limit is set to default of 83c. It seems to run cooler as it takes a little time to reach 70c, but that limit is too low to hit decent speeds. I have tried downclocking, wiped and reloaded drivers with variuos versions, running fans at 100% and replacing firmware with a couple other versions all with the same results. I have double checked the installation of the pads and they are in the same place and touching chips and heatsinks properly.

At this point I am out of options other than possible damage. If anyone has any ideas I would appreciate. them.
 
Solution
I resolved it last night. I had suspected it was the VRAM. I had installed 3mm pads that seem to make good contact, but wasn't sure they were working well enough. I put a small amount of thermal paste on both sides of the thermal pads. Now things have returned to normal. However my temps aren't any better after all of that. At least I can game again.
Oct 13, 2020
1
0
10
Have a really similar issue with GTX 1660 Ti after replacing thermal pads and paste. I now understand how hard it is to get the right pad thickness for a GPU. I went through the same process as you and re-applied different thickness Grizzly Minus 8 pads about 6 times now (its getting costly) and I'll tell you what I found out.

VRMs need to have really good contact with the pads, I used to check for the footprint of the contact and though it was decent enough (had a small visible gap made), turns out it wasn't. I was hitting Thermal Limit with 70c (remember this is only the gpu temps it doesnt show VRM or VRAM temps). Adding a second 1mm pad (2mm total) on the VRMs seems to have improved my perfomance alot. But it is very tricky to navigate the right pad thickness, the VRAMs near the GPU chip need to have a very low thickness pad (I use 0.5mm now, thinking of switching to 1mm) else the GPU wont make decent contact with the heat sink.

If anyone has good gelatinous (very flexible) pads to recommend It would help a lot, since the original pads that I removed from the GPU were very flexible and I think that was the main difference from the Grizzly Minus 8 and Arctic pads I've tried, they both are not nearly as flexible (especially the Arctic ones).

I am willing to bet that you have the same VRM overheating issue, try your luck with different pads.

Please keep me up to speed with your progress. ;)
 
Last edited:
Oct 11, 2020
2
0
10
I resolved it last night. I had suspected it was the VRAM. I had installed 3mm pads that seem to make good contact, but wasn't sure they were working well enough. I put a small amount of thermal paste on both sides of the thermal pads. Now things have returned to normal. However my temps aren't any better after all of that. At least I can game again.
 
Solution