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

SOLVED Best way to apply thermal paste

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
When the acrylic flexes, the center part lifts off the core. If you don't want to call it removing that's fine, but it is lifting off the core.
Get something flexible and play with it, you'll see exactly what I mean if you simulate what he's doing.

More politeness would be appreciated.
 
Also i am hungover and irritable right now
You still have control over your speech. I'd appreciate if you could tone down the accusations and think about what you have typed before hitting submit. If that means you wait until you aren't hung over, that is fine. The forum isn't leaving in the next six hours.
 
I think it time for a group hug :grouphug:, the main thing that appealed to me about this forum was that I saw so much respect and appreciation for others in the community. Overclockers unite! :)
 
Something else that would be appreciated is if someone who is skilled at the thin film method repeat the experiment with the various methods themselves (and film it for us).

Disclosure: I'm a thin filmer myself, but only twice.

I just can't see how having 1/4 of the processor without TIM on it is a good thing for temps, even if it is around the edges. I assume the processor heatsink material is chosen for its heat conductivity and so even the corners would have a fair wack of heat in them when the processor is running at 70C. Even if the bulk of the heat is in a penny-sized circle at the center, the edges (or corners) will be a just a few millimeters away.

For example, if you held a small copper plate and heated up the center to 70C surely the corners would still be hot enough to blister skin (60ish C).
 
I just can't see how having 1/4 of the processor without TIM on it is a good thing for temps, even if it is around the edges.
The processor die is small and placed in the center of the IHS. The farther out on the IHS you go, the less heat it will transfer. That is why the edges are not important.
 
The processor die is small and placed in the center of the IHS. The farther out on the IHS you go, the less heat it will transfer. That is why the edges are not important.

this I beg to differ.
as low as it is, surface area of contact is still important.

plus eventhough just 1'c Difference, not applying is just not optimizing your system where you could have easily done so.

The thin film method continues to be at least 1'C better than all others in around 60 mounts I tested. and true, maybe no one cares about that 1'C.

just the fact that no thermal paste can conduct heat better than the metal of the CPU top already guaranteed that heat will be found there, and since surface area is 'everything' in cooling, I will urge everyone to do it right.
 
Well after watching the video and reading..Does TIM need to cover the entire chip? I thought it did?? I'm using Artic Silver MX4. How do I make sure I put enough on, but not too little.

I'm going to use the dot method OR the thin film, spreading it across the surface of the chip with a gloved finger. I don't know how thin the film should be. I have no preference however. As this is my first time doing this.
 
Hi all,

Just saw this video, I thought you all might be interested to see it. :clap:


I am going for the pea method :thup:

Time to get pasting! :escape:
 
Hi all,

Just saw this video, I thought you all might be interested to see it. :clap:


I am going for the pea method :thup:

Time to get pasting! :escape:

another guy in the video that doesn't know how to do the film method in the video, i have no idea why he's smearing with a glove.. kinda weird.
it is near common sense that a fully covered CPU will transmit heat better, as long as you know how to film it out. :)

.... anyways. pea works for most, go for it. :)

I am sticking with thin film, gives me 1 extra 'c lower consistently.
 
Last edited:
The processor die is small and placed in the center of the IHS. The farther out on the IHS you go, the less heat it will transfer. That is why the edges are not important.

I understand that only the center is heated, and have seen those interesting pictures of people removing the tops of the ivybridge processors to replace the internal TIM.

That said, the top of the processor (IHS?) is a heat conducting piece of metal. If the center blob of it is heated to a constant 70C, what do think the temperature of that same piece of metal is at a distance of a few mm (to an edge) or a few more mm (to a corner)?

If I were a betting man, I'd bet that even the corners are at least 60C and therefore have some heat to transfer if they are in contact with some TIM.

EDIT: I'd still like to see a thin film expert repeat the experiment in the video to shed some light on the bubble controversy :)
 
another guy in the video that doesn't know how to do the film method in the video, i have no idea why he's smearing with a glove.. kinda weird.
it is near common sense that a fully covered CPU will transmit heat better, as long as you know how to film it out. :)

.... anyways. pea works for most, go for it. :)

I am sticking with thin film, gives me 1 extra 'c lower consistently.

What do you use to spread the thermal paste?
 
That said, the top of the processor (IHS?) is a heat conducting piece of metal. If the center blob of it is heated to a constant 70C, what do think the temperature of that same piece of metal is at a distance of a few mm (to an edge) or a few more mm (to a corner)?
I didn't say they don't get heated, I said they don't transfer appreciable amounts of energy.

If I were a betting man, I'd bet that even the corners are at least 60C and therefore have some heat to transfer if they are in contact with some TIM.
The problem with that logic is that temperature is not energy. It might be 60c, but that doesn't mean it is moving a lot of energy. I'd be surprised if the edges transferred more than a few watts, total. I'm not saying it wouldn't help, I'm saying it is (likely) a lot of work for little to no gain.
 
I thought we just saw a video where the film method introduces air bubbles? What I found to be most effective is a small bb size in the middle, and I put a SMALL amount of paste on all 4 corners. This gives great covereage for me everytime.
 
I thought we just saw a video where the film method introduces air bubbles? What I found to be most effective is a small bb size in the middle, and I put a SMALL amount of paste on all 4 corners. This gives great covereage for me everytime.

I think Earthdog has the "optimum" solution right here, with this method, there are no air bubbles and all the corners get covered with thermal paste :clap:
 
JoeC taught us to spread it evenly with a razor blade accross the whole surface.

Covering and spreading it evenly, to cover the entire surface.

Me?
I put a glob on and spread it with my finger making sure it has an even flow covering the entire surface with a thin coat.

Never fails.

:D
 
I know if the processor runs too hot the board or the cpu or both will act crazy and shut down..

How would you know if the chip had just barely enough to operate (ie no software crashing due to heat) but was "obviously" too hot due to poor TIM application ??
 
How would you know if the chip had just barely enough to operate (ie no software crashing due to heat) but was "obviously" too hot due to poor TIM application ??
The temperatures would be higher than normal and a reapplication of the TIM would fix the problem.
 
correct way for PK-1

apply thermal paste, thin with razor blade into a layer so thin, you can see the words on the CPU heatspread. Then add a small blob right in the very center.

Put down heatsink evenly if you do it this way. (or side first, if you are not adding the blob in the center), tighten til you cannot tighten anymore with fingers. Then use your screwdriver and add another 3/4 turn. Use springs please.

best temp Ever.
 
The temperatures would be higher than normal and a reapplication of the TIM would fix the problem.

I've never built or used this chip or used these components, so how would I know what is a good normal temperature vs a bad temperature if there is no bad behavior like software crashing and such???
 
Back