• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Thermal Interrace Material (Paste)

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

mgilbert

Registered
Joined
Jan 14, 2015
I have a five year old MSI RX-480 GamingX 8G video card. A couple of years ago, I updated the card's BIOS. Shortly thereafter, I noticed that it would stutter when drawing anything over 120 watts - it's rated as a 150 watt card. I figured the 120 watt limit had something to do with the new BIOS, so I set the limit in the driver to 120 watts and forgot about it.

A few days ago, I decided to game a little, and the card started stuttering again. When moving my character about in a game, the game would stutter hard, rhythmically, once every second or so, like clockwork. I had to lower the limit further, to 100 watts, to get smooth gameplay. I couldn't find a reason for the problem, so I pulled the card out to check for dust. The card was clean. Interestingly, using AfterBurner and HWInfo, none of the card's temperatures were high under load.

So, I removed the four screws holding the HSF assembly to the card's circuit board, very carefully disconnected one connector, and laid the card open, on my desk. There were two other connectors, but it wasn't necessary to disconnect them in order to lay flat the HSF and circuit board assemblies. The TIM looked fine, but I cleaned it away from both the HSF and the GPU chip, using a soft cloth, isopropyl alcohol, and swabs. I replaced the paste with fresh Prolima-Tech PK-3, and reassembled the card.

And that solved the problem... The card now works flawlessly at full throttle again, with no stuttering. So, even though the TIM was not dried out, and temperatures looked fine when gaming, the TIM was not doing its job.

So, if you're having issues with your GPU under load, and it's been a long time since you replaced its thermal paste, replacing it might be worth a try. Doing so is a delicate procedure, best left to an experienced electronics/computer technician. If you don't know what you're doing, you can destroy your card, so proceed at your own risk. It's worth noting that some cards are harder to disassemble than others, so do some research before you begin.

I've never had to replace thermal paste on a CPU, so it never occurred to me that the problem with my GPU was indeed the TIM. I'm posting this in hopes that telling my story will help others.
 
Interesting. I wonder if it was something else (during the breakdown/reinstall process) and that changing the paste was just a coincidence. If the paste wasn't doing its job, then temps would be high, ya know? So the paste was working.

Thanks for sharing!
 
Interesting. I wonder if it was something else (during the breakdown/reinstall process) and that changing the paste was just a coincidence. If the paste wasn't doing its job, then temps would be high, ya know? So the paste was working.

Thanks for sharing!

Good point... It could have been that an area of the GPU chip that isn't monitored was overheating. Or, perhaps there was a loose connection somewhere, and pulling the card, disassembling, reassembling, and reinstalling the card resolved the issue. We'll never know for sure. But it's fixed, which is good, considering the current prices of GPUs. I paid $300 for this GPU, and until I can buy a card that's twice as fast for the same money, I am not upgrading. If that's "never", it'll just have to be. I'm not paying $500 for a midrange card, ever.
 
Back