So I currently have an RX Vega 56 Asus ROG STRIX. I know that probably based on these temps you see you’ll say it is normal. But here, I was using the AMD self stress tester for around 3 minutes. However, in God of War 2018 GPU temperatures go up to 80 degrees and the memory up to 83 or 85 max but I’m not so sure about the VRM temperatures. Since God of War doesn’t let you really see properly. And the only reason I’m worried about the VRM temps is because there is a specific part in the game where the game runs at 80-90 FPS then outta nowhere drops to 55-60 as well as the GPU usage from 99 to 60%. So I thought it was probably the GPU throttling it self to keep cool, idk. I would really appreciate what you guys think it is since this GPU has never gotten a thermal paste replacement nor thermal pad replacement. Also if it can damage itself or not.
And yes, my Vega 56 is a bit undervolted by 50 mV in p7,6,5,4 and I can’t really undervolt it any lower than that since I got a pretty bad chip it’ll crash under that. So, basically at 1150mV.
2 Comments
Yeah it’s safe, VRM mosfets can take the heat, but I think it can be improved. From [looking at a teardown](https://www.reddit.com/r/Amd/comments/8zn0yh/strix_vega_56_disassembly_and_thermal_repasting/), they have a pretty thick (2.5mm?) pad on the VRM, which isn’t great for cooling. I’d recommend U6 Pro thermal putty (aliexpress), since it can fill around the mosfets and pull more heat out of the PCB. You could also replace the paste on the core with PTM7950, if you want to go nuts.
Might be time for a re-paste and a re-padding. Be very gentle around the GPU, the memory dies easily. My Vega 64 had worse temps with the reference blower cooler. It came to life with a waterblock. Good luck finding one as it took me a minute to secure one back when that card was still in production.