Very high 3080 VRAM temps, anyway to fix it?

You could also just pull the core clock down as far as afterburner allows and drop power limit down to until vram temps are good. Selct a static fan speed that you are happy with and save it as your mining profile.

Your hash rate will be lower, but so will your temps.
 
You could also just pull the core clock down as far as afterburner allows and drop power limit down to until vram temps are good. Selct a static fan speed that you are happy with and save it as your mining profile.

Your hash rate will be lower, but so will your temps.
In the past I have lowered the core clock and set a 70% TDP, however it seems to only bring down the temperatures by 1 or 2 degrees
 
In the past I have lowered the core clock and set a 70% TDP, however it seems to only bring down the temperatures by 1 or 2 degrees

In order to get the card to pull power from the vram you have to go lower on the power limit with the core clock bottomed out.

50% got me to high 80's on vram.

The vram only needs about 100w to get hot and with the GPU not doing much, a 70% power limit leaves plenty of power available to push to the vram.
 
Wonder if you just drop vram clock speed will cure the overheating without touching the core. Alternatively drop tdp and core speed
 
I have a 3090 MSI Gaming Trio X that runs very hot on the vram - I've just taken the side off my case now and I can get it to sit about 104c when mining.

Before I "discovered" mining a few months ago I had actually bought the card to game hard on my Samsung Odyssey G9, however I found that Assassin's Creed Odyssey (there is some irony there) would sometimes completely crash and reset my machine. I'm now wondering if this was vram temp related as my case is pretty tight without the side on. Seemed to just happen when I hit the menu button to bring up the map screen, very strange.

I think I'll be investing in a bigger better case - I have the 3090, i9 9900k crammed into a Fractal Design Define Mini C just now.
 
I have a 3090 MSI Gaming Trio X that runs very hot on the vram - I've just taken the side off my case now and I can get it to sit about 104c when mining.

Before I "discovered" mining a few months ago I had actually bought the card to game hard on my Samsung Odyssey G9, however I found that Assassin's Creed Odyssey (there is some irony there) would sometimes completely crash and reset my machine. I'm now wondering if this was vram temp related as my case is pretty tight without the side on. Seemed to just happen when I hit the menu button to bring up the map screen, very strange.

I think I'll be investing in a bigger better case - I have the 3090, i9 9900k crammed into a Fractal Design Define Mini C just now.
All these AIB cards vent out of the top and bottom. So if the top of the card is close to the side panel then it could be restricting airflow and since the air gets pushed against the side panel, some of it will go back down and get pulled in by the fans.
 
As there is a lot of interest now about this topic, I just want to add that a few days ago I had an EVGA RTX 3080 XC3 black that was mining at a very mild 53% power limit and 80MHs ish. The card had been mining at more or less that TDP since I got it. Then it just simply died. No display. 3 months old. Fortunately the retailer I got it from is absolutely brilliant because the next day, without returning the card which they even came to collect for free, they sent me an EVGA RTX 3080 FTW3 Ultra gaming as a replacement, which is a more expensive, power hungry card. It works perfectly so it was definately the GPU that died. I'm now very dubious about mining. Maybe it was luck of the draw, who knows. I always had my eye on those temps and the VRAM never exceeded 96c. And 82c while mining.
 
As there is a lot of interest now about this topic, I just want to add that a few days ago I had an EVGA RTX 3080 XC3 black that was mining at a very mild 53% power limit and 80MHs ish. The card had been mining at more or less that TDP since I got it. Then it just simply died. No display. 3 months old. Fortunately the retailer I got it from is absolutely brilliant because the next day, without returning the card which they even came to collect for free, they sent me an EVGA RTX 3080 FTW3 Ultra gaming as a replacement, which is a more expensive, power hungry card. It works perfectly so it was definately the GPU that died. I'm now very dubious about mining. Maybe it was luck of the draw, who knows. I always had my eye on those temps and the VRAM never exceeded 96c. And 82c while mining.
Not sure which retailer would do that as most like to atleast inspect and test a card before sending a replacement let alone the replacement being a model 2 tiers up.
 
BJF3Cv8.jpeg
Not sure which retailer would do that as most like to atleast inspect and test a card before sending a replacement let alone the replacement being a model 2 tiers up.
I was just as surprised as you mate, they were brilliant. The FTW3 is huge compared to the XC3 Black as well.
SIZQCiM
 
My 3080FE mem temps were hitting 110c with 80% fan +800 mem mining. I swapped the thermal pads for Thermalright Odyssey 1.5mm and temps dropped to 84c 55% fan and +1300 mem. Gained 8MH/s so now 99MH/s. I used a plastic card to carefully remove the OEM pads and kept them so if I ever need to RMA I can put them back.

The FE has no anti-tamper stickers, just be careful not to mark any screws and they will never know you were in there. The thermal paste they use looks exactly like Thermal Grizzly Kryonaut
 
Nice, was it a UK retailer?
It was yes. I mean I still had to wait for the black XC3 to be available for pre order in the first place like everyone else. That was back in October. I was lucky. Got the card right on new years :). They have a 48 hour replacement policy on marked products (like my black XC3) which guarantees a 48hr replacement if its under warranty.
 
Last edited:
I am an overclocking idiot, but can't you just under-clock the cards memory?
Yes, You can but you'd just be wasting your time. This was my first thought before I knew more about what I was doing. Thinking, well obviously if I'm overclocking the memory, the memory will just run hotter, right?
But actually, even lowering the memory to -502 (The lowest MSI Afterburner will possibly go) all you end up doing is maybe losing a few degrees but your hash rate will absolutely tank. I mean If that works for you and your happy
with the tradeoff from 94MH/s ish hashrate on a 3080 to about 76MH/s for the gain of about 8c at the absolute very best even with a desk fan blowing on it, then sure go for it.
But running Nicehash's recommended settings and using an open air case and deskfan blowing like this gives me 94MH/s.... VRAM right now is 78c, but the desk fan is on it's fastest setting. Was looking at maybe an air con unit to help as summer approaches.

Read here for Nicehash's optimum settings (Recommend afterburner over EVGA Precision X1 as I seem to get lower hash rate with the same settings...)
NVIDIA RTX 3000 series Overclocking Guide to Increase Mining Profits | NiceHash
And this is how I have it set up, only temporarily though. When gaming I just put the fan aside and panel on. I clean it weekly. Air conditioning is next as having the windows open all the time means dust...hair....bacteria....grime....mould... all sorts.
8GML080.jpg
 
I recently managed to get a 3080 FE for my new pc, however I have noticed very high temperatures for the GDDR6X VRAM with it reaching up to 106 degrees Celsius, I primarily use it for gaming but I want to mine on it while im not playing games to make some money back on the purchase however I am concerned such high temps will damage the card in the long run so wanted to check if anyone had a practical solution short of watercooling it?
I have already attempted undervolting without the desired results, it only ends up reducing the core temp however that is not much of an issue as that tends to never go over 65-70 anyway.

Hi, i added an old harddrive water cooler on the backplate. Now everything is cool and silent. MJT 78°C, GPU 45°C mining 24/7.

https://ibb.co/zHKLX5S
 
This does not seem right to me as these 3-Series cards get very hot vram and they were not made for mining, and to run at 105-110c 24hrs , 7days a week, all year round, will definitely cause issues in the long run and maybe kill off the card?

In gaming you're likely doing 70-95c max couple of hours every other day.
 
Back
Top Bottom