That's basically what I meant. You only expressed it in more technical terms. :)

No problem.
:D

Looks like fate-is-one-edge already answered your main question, but I'm going to second his recommendation to cap your FPS. If your GPU is only working at, say, 80%, that'll keep temperatures nice and cool. I'm running Spec Ops: The Line on my 940MX right now, and capping it at 30FPS (rather than allowing it to hover between 40 and 50FPS like it usually does) has shaved between 7 and 8 degrees from the peak temperature. Granted, this is on a laptop, where space is more constrained, but you should still see a considerable difference.

The card is not even a year old, you were just wasting your time, you don't have to replace thermal paste once a year, maybe once every 2-3 years and only if you notice a problem with temperatures.

As for other advice people recommend to put much more thermal paste on GPUs compared to CPUs. So if you put the small amount that is recommended for CPUs, you might have not used enough. But considering that you are running the game at an unlocked framerate and the GPU is stressed out and running at 96% performance the temperatures look reasonable. Use Riva Tuner Statistics Server to lock framerate to 60 FPS if you want to give your GPU some breathing room.

The factory applied paste is usually not of high quality (like Noctua NT-H1, for example) and sometimes it's even poorly applied. I've seen one case in which a piece of plastic sheet was stuck between the GPU die and the radiator, covering about half of the die.


OP, if the advices already posted here won't work and you're not comfortable with the current temps, then you could try to undervolt the GPU a bit and lower the frequencies.

There are some good answers here but the question is, what temps did your GPU run before the thermal paste change?

I am not familiar with that model but 80º is a bit high for a 2070, since Nvidia claim max temp is 89º, I would expect something around 70 to 75º while gaming.
Pascal max temps on Nvidia website are 94º, not 89º.

To add some further info to what's already posted, check again if your cooler is properly seated on the GPU, otherwise, it can make temps sky rocket.

Here you go! :D

Like this?
This was a Sony laptop that was quite noisy from the fan... Guess why...

Yes! :D Was that your laptop?

No, I had it here for repair. Was Sony brand, cannot recall the model but had a i5-3337u (I know because I still have the cinebench score for that cpu saved) and the GPU was a Nvidia 940mx or something like that...

Edit: One of those cases where the GPU vastly outperforms the CPU power...

Are you running with unlocked framerate? If yes, around 80º is normal.

I clearly remember Arctic Silver 5 not being recommended for graphics cards and in its instructions it says that you need to heat it up and cool it down several times (so turn off/put computer to sleep) for the best effect. Alright, so no need to remove it when it is there already. Not having one of those GPUs I cannot tell whether the temperature is as it should be.

Searching around I stumbled upon what must be the OP's post in a technical forum :-) Good to see you solved the problem! Indeed the best result comes from the correct amount of TIM applied with no air bubbles and a good, even pressure from the cooler on top.

Too much is bad, too little is bad. Carefully remove ANY SPILL as that TIM is conductive. You do not want a short on such an expensive GPU.

Can I ask about that plastic bit? I have one on top of my graphics cards and thought it was just a leftover that the builder put in.So,the question is are they important and protect something? And if they are,why don't they secure them better?

Okay,Fate-is-one-edge has cleared up the little mystery,thanks.