What's the best way to apply Artic 5?

artic silver 5 website

i've found the best way to apply AS5 is to apply a line the length of the chip directly in the center and use the edge of a credit card to spread it until it's just thin enough you can see through it. Then apply AS5 to the HS plate in the same fashion covering the whole HS plate in that thin almost transparent layer.

Once you get skilled enough at the application you really only need to apply the AS5 to the HS, if you aren't making good contact during the mounting then it doesn't matter how much AS5 is on either the cpu or HS.

No, follow the instructions on the website exactly. There's too much room for error otherwise. You hardly need to use any of the stuff at all, too much is in some cases much, much worse than not enough. There's really no "trick" to applying it.

I would say it would be ok as long as you monitor your temps and don't go overboard. What are you using for CPU cooling Stock or do you have an aftermarket cooler?

Their AS5 instructions somewhat baffle me(especially before reading the instructions for the single core with heat spreader with decent pictures. Their instructions for a dual core just say a "thin line" in the right direction, and instead of a picture like all other types of CPUs, they have a little red line showing you which gives no idea how how thick the line should be, or how much should be on there. "thin line" can be in how thick it is or also how wide it is, but they don't say what they mean so I'm quite confused by it all. Does anyone know where any pictures are for AS5 pre-applied to a dualcore? I'd like to get an idea of how much is used.

Thanks you!

You just need enough AS to completely cover the CPU with a thin layer, like a paper sheet thick. You can use several things to spread, a credit card is a good tool to do it. do not leave any holes, just cover completely, and carefully. That should take you like 5 minutes or so to complete.

About the quantity, I would use a thin line (like 1mm wide) in the center of the CPU, starting and ending 5mm away of the edges. That will be more than enough to cover the CPU completely.

Sure you don't need to, but you need to have a very, very good pulse to do this (something personally I don't have). If you want teh paste to be spread covering completely the CPU you should do it by yourself. Almost every builder will tell you that.

Forgot to mention, it's not important for the entire heat spreader to have AS5 on it. Only the interface between the internal dies, the heat spreader, and the HS need to be covered. The "spread it yourself" theory is a hold-over from the good ol' days before integrated heat spreaders, when it was critically important to make sure the exposed core got good contact or parts of it would die. Those days are gone.

THank you for the replies! They help quite a bit... like ti was mentioned... I haven't applied TIM since the days where you did it on the die and made sure to get a thin layer over all of it. I order some good X23 TIM and a Tuniq Tower 120 and will use the apply a small amount and let the HSF pressure spread it out method.

Thanks!

Haha, sometimes is important.... Anyway, I'm not saying that the pressure method is incorrect, in fact AS signals that one as the "recommended" way to apply it, just I say I feel more comfortable spreading the paste myself. When I do it I'm totally sure that the paste is applied totally even across the CPU. And a lot of people also use that way.

You can go the way you want. The important thing is that you are happy with the results.

One of the advantages of spreading yourself is that you minimize the air bubbles. And if you spread a very thin layer of paste it won't go nowhere. Even more, I'm not using AS5 now, because I switched to Coollaboratory Liquid Pro (even better than AS5, need to be a lot more careful using it tho) and it is almost (if not all) pure liquid metal, and the way to apply it is also to spread it all across the CPU.

There is A LOT more "risk" with that product (because of its conducting characteristics)... But the thing is, if you know how to use it, you will be fine, and will achieve excellent results.

Anyway, you are 100% right. If somebody is has no experience doing this kind of things, it's maybe better to be on the "safer" side. But AS5 is a safe product in the first place.

Apply an amount about the size of an uncooked grain of rice. I usually just spread the stuff out with my fingertip until its good and even all the way around.

It is actually important that you DO spread the paste out prior to applying the heatsink. Especially on Intel systems - due to the way the heatsinks are mounted, your paste will end up squished to one side of the ISH only.

I haven't found this to be the case -- the compound tends to be spread out uniformly from the center in a circular pattern (I am assuming you are talking about LGA775 systems).

Yea, I like the CLP stuff too, it makes me kinda nervous though. When you spread it you're introducing the risk of air-bubbles, it's pretty much impossible to get them when you just squish the stuff down with the heat sink, but if you're careful you won't. Resistivity isn't the only concern, even if a material is non-conductive it will act like a capacitator if it migrates down in between your pins, and that's just as bad. Sooo many times I've gotten random shut-down calls on homebuilts, and when I take a look, sure enough, these people have used the whole damn tube of AS5 to mount their heat sink. High temps, weird behavior, CPU death. Much easier for most to just drop a dab on there and let it be!

What makes you say this?

From what most believe, after a burn-in period it works better and your temps sometime drop by up to 5*C.

Ok the credit card method was taught to me by a microprocessor engineer, so i think it's a safe bet on it being a good method.

Yes, you do want to apply the AS5 to both the HS surface and the core/heat spreader on the cpu. They should have a layer just thin enough you can almost see through...almost.

You never...never ever spread AS5 with your finger, bad idea.

Another thing i do to prep the HS surface is coat it a layer of AS5, and then use a bleached white coffee filter and rub the AS5 in until it's clean before applying the layer for the cpu. Also if using a credit card, (which i again stress is the best method for having a thin even layer) is use the coffee filter to wipe off the card. The reason for using the coffee filter is that it doesn't have any lint or material that will stick to the surface ensuring an even layer of thermal grease.

No disrespect to these guys but very few other thermal compound reviews agree with their findings. By itself that means nothing. However when you combine the result of multiple reviews you do see a pattern which tends identify outliers. Maybe the others have something to gain or flawed testing but it is something to keep in mind.


I personally have found no way to get Arctic Silver Ceramique to outperform Zalman ZM-STG1 or AS5 with my tuniq tower. Hell even once I tried it with no compound. It actual performed quite well. At idle and under a quick load test performed the same a using the Zalman. I didn't do an extended load test.

As for the burn-in I have seen little or no difference

lol yeah i have done the same.

some think that will kill a cpu, it wont. it will run hotter but
it wont melt

Ok, here it is, rather entertaining stuff. Enjoy!

TIM Shootout

cool i havent seen that in a long time, but i do remember it.

So if you improved your air flow in the case... it should do quite well compared to the 9500, correct?

No product on the market has a significantly lower thermal conductivity than the "average" heat transfer compound. Thermal interface materials, are, as a given, fairly poor when it comes to thermal conductivity (due to the inclusion of various oils to prevent drying out, silicone oil being the most common).

What I was getting at is that people seem to think the barrier between the heat sink and the CPU is highly significant when it comes to dissipating the heat on the CPU, and this is simply not the case. The biggest "bottleneck" is between the HS and the ambient air. That's why when you look at a bunch of thermal compounds the difference is only a few C or fractions thereof, whereas the difference between air and water is often in the 10's of C. Sorry if my wording wasn't clear.

Hmmm, I knew someone was gonna bring this up....

Coollaboratory Liquid pro isn't a thermal interface material, at least not in the traditional sense. It's freakin' pure liquid metal. It doesn't have the oils that almost all the other do that act as thermal insulators. So it truly is in a class by itself. It's interesting to note that the amount of improvement between this and traditional compounds ranges from 2 to 6C under load, much less than a switch to liquid cooling.

But I'll still give you a "D'oh, you got me."

Respect.

Go ahead an take it off, use a paper coffee filter and carefully wipe away the excess material, buff it like you're waxing a car. Then reapply a small amount as instructed. Pay attention to the sides of the heat sink and cpu heat spreader, make sure you don't have any drippage down there. Don't worry about alcohol unless some got where it wasn't supposed to be. Good luck.