Quality thermal paste, how big of a difference can it make?

Here is the situation, the 9800M gs in my Asus G50vt escalated to high 80 C(close to 90 sometimes) lately while gaming, perhaps it's because of high ambient temp. But still I don't like it at all, I have the factory thermal paste replaced with some thermal paste(don't remember the brand, but it's not AS5 or IC diamond, most likely some generic paste) from my friend many months ago.

I've heard many good things of IC diamond-7, my questions is will it make a difference if I replaced my current paste with it? I am NOT aiming for +5 C decrease in temp, even 3-4 C decrease will justify the purchase for me. Please post your experience with IC diamond-7 and its performance!!

Answers/advices will be appreciated!! Thanks!

 

Application is more important than paste (except for the cheap junk OEMs use). You shouldn't have a difference of more than 1-2 degrees C between different pastes if applied properly.

 

I may be wrong, but that temperature does not seem high to me (while in gaming mode).

As to whether you will see a decrease, it depends on the quality of the paste job the machine has now and how good your abilities are.

Others will answer better than me, but I think you'd be wasting your time given that your current temps are very reasonable.

When your friend first did that repaste job, what were the temps then?

 

I agree, 80 celcius is something I can only dream of (my gpu reaches 92-94 celcius). I wouldn't worry about those temps at all.

 

It's in the high 80 range, meaning it's close to 90. I did the paste job myself, but the paste came from my friend. My question is will quality paste like IC-diamond-7 will make a difference in temp drop?

 

I have to agree with notyou, the expertise with which is applied is more important than the type used.

So, when you initially did that paste job, what were the temperatures then (before/after)?

 

It was probably about 1 year ago, a few month after the purchase. The GPU temp stayed under 85 C all the time back then. It seems to me that the GPU temp keeps escalating over the time. Will the paste I applied lose its functionality over the time?

 

The paste does deteriorate afaik.

Have you been keeping the fans/vents clean since then? Have you used air to clean the notebook out?

Maybe thats all it needs.

With what you stated about your temps, it seems like you could achieve your ~5 degree reduction in temps if you re-pasted the gpu, but remember to take into account that a year ago you might not have been using a game that was pushing the gpu as hard as you are currently too.

For what its worth, I too have heard very good things about ICD-7. But again, a competently applied paste is worlds better than the best compound applied sloppily (or by an inexperienced user, like me ).

See:
http://forum.notebookreview.com/asus-gaming-notebook-forum/506044-chastitys-repaste-experience.html

 

I clean the inside of the notebook regularly with compressed air. If ICD-7 can indeed make a difference of 5 C compared to whatever paste I applied before, then I buy it. It's true that the game I used to play might not stress the GPU as much as I did now with current games, but still I am looking for any means to improve the temp.

 

Regular cleaning with CA - great.

I didn't say ICD-7 by itself will make a difference.

If the games are stressing the gpu more, then you'll eventually run into limits (of diminishing returns) of how cool it will run 'during' those games.

Fine to keep looking for 'any means' of improving temps - but keep in mind that a repaste at this time may not be enough?

Have you tried a notebook cooler?

 

I have a Zalman NC2000 underneath the laptop when I game, it does help lowering the overall temp quite a bit, but just not enough to keep the GPU temp under 85C. I even drilled lots of small holes on the backpanel to improve air circulation and it also proved to be very helpful. So far the only mean to lower the temp I can think of is to repaste the GPU, that's why I am so counting on a quality thermal paste.

 

Make sure to check the difference in height between the heatsink and GPU, since sometimes you'll need a shim if there is a significant gap. As for pastes, the differences aren't great between the top recommended ones (ie. AS5, IC7, etc.). If one is significantly cheaper than the other, I'd just go for that.

 

Actually if there is a gap or loose fitting of the HS to the chip ICD is a much better choice. ICD actually will hold up longer setteling to a dried out thin pad between the HS and chip. Because the diamond is such a great heat conducter the thicker interface is fine.

After 1 year on the CPU and northbridge on my P7805 I replaced the P9600 cpu. When pulled apart the TIM was making 100% contact without dry out, cracking or boilout. Personally I'd never use anything other than ICD again, current state of the art that is.

On my CPU I got about a 3C drop with ICD from MX3. My GPU runs very cool so I hadn't tried it there.......

 

While yes a more competent thermal paste application is more important then the thermal paste used, when a competent application of say AS5 and a competent application of say Shin Etsu x23-7783D are put against each other. The Shin Etsu will allow for better heat transfer, thus lower temperatures.

I'd recommend:
IC Diamond 7
Shin Etsu x23-7783D
AS5
Or OCZ Freeze

Freeze/AS5 will give similar results of about 1c-3c cooler then generic junk.
IC DIamond 7 will yield 1-2c cooler then AS5, and Shin Etsu 1-2c cooler IC Diamond 7.