Overclock.net banner

1 - 9 of 9 Posts

·
Systems Engineer
Joined
·
567 Posts
Discussion Starter · #1 ·
http://www.youtube.com/watch?v=EyXLu1Ms-q4&list=HL1361923680

I did a spread of sorts, and also feel like I put on too much, and now I'm noticing quite the temperature difference in cores. Namely core 1 and 3. Right now their sitting about 10-13c apart from each other, but I caught them spike the other day like this:



So, I grabbed some Antec 7 from work to use, but does it sound/look like I have air bubbles?
 

·
Registered
Joined
·
656 Posts
Well, you'll get about 15 or so ways to apply TIM, so here's mine.
Use a dab about the size of a grain of rice. Then spread it out over the entire IHS. And yes, it will be very thin.

Proof it works, max temp 72C at 4.9GHz:

 

·
Premium Member
Joined
·
10,885 Posts
The generally accepted way is to put a blob or line in the center of the CPU, and then just press the heatsink on and attach it. The force of the heatsink retention will force the paste out over the CPU without bubbles forming.

But you can't judge using idle temps because the cores are never actually idle, there are always background tasks that are going to be putting load on individual cores. Check it under load instead - there will still be a split it just won't be as big (usually 5-8 degrees).
 

·
Registered
Joined
·
447 Posts
Quote:
Originally Posted by Forceman View Post

The generally accepted way is to put a blob or line in the center of the CPU, and then just press the heatsink on and attach it. The force of the heatsink retention will force the paste out over the CPU without bubbles forming.

But you can't judge using idle temps because the cores are never actually idle, there are always background tasks that are going to be putting load on individual cores. Check it under load instead - there will still be a split it just won't be as big (usually 5-8 degrees).
This is the method I use with the build I have now. I remember when replacing my stock cooler on my c2d I used this method as well. Just put a little bit of that sweet sour cream in the middle and ram it with your cooler nice and slowly cream it all around. Tighten that thing in nice and deeply and be proud of your creation. A nicely working cooler that runs cool.
 

·
Registered
Joined
·
299 Posts
Quote:
Originally Posted by OS-Wiz View Post

Well, you'll get about 15 or so ways to apply TIM, so here's mine.
Use a dab about the size of a grain of rice. Then spread it out over the entire IHS. And yes, it will be very thin.

Proof it works, max temp 72C at 4.9GHz:

This is generally accepted as the way NOT to apply thermal paste. Luckily thermal paste is more forgiving than people think, but the above has been shown repeatedly to be the worst. That he has decent temps is irrelevant IMO since this issue has a plethora of data out there contradicting him. Not to mention, in my younger days I experimented quite a bit on my own since there was less data out there and found spreading to be the worst as well. It is the most likely method to trap air which is your worst enemy. One small spot in the middle is the best general method. Some people swear by lines for better core coverage but it is really rather meaningless since the differences in results are usually undetectable.

Bottom line: don't spread it, don't use to much, check your temps, don't sweat it.
 

·
Registered
Joined
·
656 Posts
Quote:
Originally Posted by Backwoods166 View Post

This is generally accepted as the way NOT to apply thermal paste. Luckily thermal paste is more forgiving than people think, but the above has been shown repeatedly to be the worst. That he has decent temps is irrelevant IMO since this issue has a plethora of data out there contradicting him. Not to mention, in my younger days I experimented quite a bit on my own since there was less data out there and found spreading to be the worst as well. It is the most likely method to trap air which is your worst enemy. One small spot in the middle is the best general method. Some people swear by lines for better core coverage but it is really rather meaningless since the differences in results are usually undetectable.

Bottom line: don't spread it, don't use to much, check your temps, don't sweat it.
EDIT: I made an inappropriate response, please disregard.
 

·
Premium Member
Joined
·
3,821 Posts
That video is correct. I saw the original version with talking and no bad music! But it is the way to do it. I big rice is about right.

This does NOT work for Direct Touch coolers. The ones with heat pipes that touch the CPU. But any flat will do this. And remember, it will spread a bit more once it heats up.
 
1 - 9 of 9 Posts
Top