Thermal Interface Material (TIM) best practices

Hi folks .. I have yet to complete my build as I am perplexed. at all the techniques I have found on the Internet and YouTube. I am installing a 2011.3 socket with a Hyper 212 Evo cooler.

I'm leaning towards doing it as the manufacturers website for arctic silver 5-- that is they recommend tinting the cooler and a line on the cpu

I also found this demo on the matter:

https://www.pugetsystems.com/labs/articles/Thermal-Paste-Application-Techniques-170/


I was hoping to get feedback from any rig vets here if possible... I found some older threads here but was also wondering if some TIMs are better than others and obviously I have a 6 core processor that may run hotter than older cpus. Maybe there isn't a "best practice"?


Just for laughs I found this video... the Kentucky way of applying things:

http://www.google.com/url?sa=t&sour...4MTZz0TEh2jUgddjQ&sig2=MOhZ_Js6Y52X_fvLY6G4PA

The difference between the best and worst TIM is 3C. Use what you have.

As far as application methods...everyone is going to have their method and swear by it. The reality is, I do not believe any empirical testing was done to see which is best. However I leave you with this video:
That said/watched... I use the tiny pea size in the middle.

EDIT: Your puget link was the most empirical I have seen.. not much variance at all. The goal is to get good coverage and no air bubbles. Blob or X seems to be it.

The 5 dots works well too but I use the pea method.

Even the Puget chart shows only a 2C difference in methods. Just make sure you apply the correct amount and get a good seat.

I just put some on the cpu use my finger or what I have handy to spread it around half way even.

Not really a good way to do it...

Direct from the people who make them for a living, really can't argue with it:
http://www.intel.com/content/www/us/en/support/processors/000005576.html

EarthDog said:

Not really a good way to do it...

Click to expand...

To tell the truth, I never could tell if it really made any difference at all on how you put it on. The amount and how good the sink was seem to matter more...to me.

Feel free to watch the video and see why the prespread method is supposedly inferior.

Again, the goal is to get the entire IHS (which the spread method does) with the least amount of air bubbles. As you can see from the video, that method introduces air bubbles into the mix.

As the others have said don't sweat the details to much.

One thing I always point out though is if you are using a paste that is conductive you want to be careful not to use to much that you have it squeeze out and potentialy short something. With that being said it takes a considerable amount of excess paste to really get to the point were you are shorting something.

Also, dont use your finger as it introduces oils into the tim (or use a gloved finger if you must use a finger).

Since 90% of the processors I've ever used where de-lidded, the Spread method is the only way.

Under the IHS plate, the TIM is spread. So to follow this tradition, with or without the IHS plate, I've come accustomed to spread method.

Amount of TIM should be determined by how concave the IHS plate may be. If you lap the cooler AND IHS plate, you can use a smaller amount of thermal paste because the mating surfaces will be pretty close to flat.

If you do a good Lapping job, the cooler on the cpu should practically suction cup together (without paste). Then you know there's a good flat surface contact.

____________

A note on mounting pressure. Be sure there's plenty. Any cooler fastened to a cpu should be difficult to move when clamped. If it moves easily, there's not enough clamping force to effectively transfer heat.