Topic: R9 380 - overheating vram? - Need help (Read 485 times)

You can not believe water is wet, but it's still gonna be that way no matter what you think

The lack of a thermal connection between the heat sink and RAM is often more of an issue than VRM cooling.
http://nerdralph.blogspot.ca/2017/01/hot-video-cards.html

I have some cards as you mentioned - and VRM is hot too - past summer i saw about 104 in peak heat, i set 65watt fun set directly on them and that help a little.

Thanks guys.

The highest temp that I saw was about 99C when playing Crysis.

About this thermopads, where I shuld put them?:
https://www.techpowerup.com/reviews/ASUS/R9_380X_Strix/images/front.jpg

I think VRAM could handle about 125°C. You could buy heatsink for them, on ebay or aliexpress they are pretty cheap and thermal paste. Check for memory errors if they are hot it start getting memory errors or it happens to my RX 480 Asus Dual.

First things first, VRMs apparenty can handle much higher temps than the core itself so 90°C is fine by some. But I don't believe that.

I used to have a bunch of Asus GTX 780 Ti cards with ~91°C VRM temps which I could get down to about ~84°C. What I did was I bought a couple of thermal pads and put them on the VRMs so they were connected to the heatsinks through the pads, because otherwise there were no connection by default. I still think that specific Asus design was faulty.

I had Gigabyte 780 Ti's as well, but while they had no temperature sensor in their VRM, as per a cheap IR gun they were much cooler as they made contact to the heatsink by design.

So if it worries you, based on the design, I would buy some cheap thermal pads to put between the VRM and the heatsink.

Hello guys.

I'm new here, but I'm reading this forum from couple weeks. Because of this forum I started my own journey into Crypto Currencies world.

I bought used (but it still got 1,5y warranty) R9 380 Strix 2gb couple weeks ago and i started mining (couple hours a day and a few nights). I treat this mainly as a hobby, mining only Zcash (cards produce about 180 sols/s at stock clocks) and in the meantime learn about awesomeness of cryptos in general. I used MSI Afterburner as a hardware monitor and i settled a little more aggressive fan curve. Reading was fine, car under full load (mining or gaming in general) run at 65C at 40% fan speed. But I was curious what could be power draw. So i downloaded GPU-Z which didnt show anything different than MSI) and HWinfo. I was totally shocked when I saw 87C on Vram sensor. I immediately turn down miner, speed up fans and started thinking what should I do.
Vram is idling at 72C, during light task (ie watching movie). Is this could be a failure in sensor (vrams in theory should produce low heat)? My case is Phanteks Enthoo Pro, so airflow souldnt be a problem, PSU is Corsair RM650x.

Can you guys help me out? I dont think that mining/gaming on this card is safe in general (and I dont want to run fans at full speed all the time).