• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Question about applying thermal paste.

Azureth

Azureth

Supreme [H]ardness
Joined
Feb 29, 2008
Messages
5,323
I have Arctic Silver 5 and am going to use it on my Intel Xeon E3110 with my ARCTIC COOLING Freezer 7 Pro HS. I've watched some videos online but they've all said different things, some say to apply thermal paste to both your processor and your HS, some say to apply it and spread it around, while some say to just put a dab in the center of your processor and leave it at that.

So my question is, what is the proper way to do it? Does what kind of paste and HS, or processor I use matter?
 
I follow these 2 rules and have never had a problem:

1. If CPU and HS are lapped, apply a rice grain sized dot in the middle of the CPU IHS and plop the HSF down onto it, give it a 2-3 degree turn clockwise then counter clockwise and lock it down. Heat will do the rest. Just make sure to apply even pressure when turning the HSF.

2. If CPU and HS are not lapped, use the line method. This will ensure a sufficient amount of AS5 is applied to fill in imperfections in the HS base and IHS of the CPU.
 
HydroBudz said:
I follow these 2 rules and have never had a problem:

1. If CPU and HS are lapped, apply a rice grain sized dot in the middle of the CPU IHS and plop the HSF down onto it, give it a 2-3 degree turn clockwise then counter clockwise and lock it down. Heat will do the rest. Just make sure to apply even pressure when turning the HSF.

2. If CPU and HS are not lapped, use the line method. This will ensure a sufficient amount of AS5 is applied to fill in imperfections in the HS base and IHS of the CPU.
Click to expand...

I'd recommend using the line method in either case. C2D/C2Q chips are different than older generation chips... back in the day, the rice method was a good one, but it's just not the case today IMO.
 
I also say the line in any case.
Vertical for duals, Horizontal for quads (Unless they've changed...). I only do dots on AMD processors.

Then again, I'm not a thermal paste professional.
 
I was wondering, it states that when used once you should put another cap on it and leave it in a cold place like the refrigerator or it will lose its effectiveness, is that true? The AS5 I have was used on another computer I had built about 3 weeks ago and since then it's been in a bag with other stuff in storage. Is that really a problem?
 
just spread it on and slap the heatsink on it. worrying about crap like this is ridiculous.
 
UtopiA said:
Spreading actually takes longer than the line method...
:p
Click to expand...

Let me be more PERFECTLY clear then: It doesn't matter. It's stupid to have threads about it. The people who worry about this kind of thing are the same people who preached for years about how 'horrible' ram at 2T was, how sandra's memory benches are incredibly meaningful, etc. etc. All this little minutiae crap is meaningless.

Your car runs a hell of a lot hotter than your PC, and I don't see people out there making special lines and dots with the various fluids in their cars.
 
retrorocketz said:
Let me be more PERFECTLY clear then: It doesn't matter. It's stupid to have threads about it. The people who worry about this kind of thing are the same people who preached for years about how 'horrible' ram at 2T was, how sandra's memory benches are incredibly meaningful, etc. etc. All this little minutiae crap is meaningless.

Your car runs a hell of a lot hotter than your PC, and I don't see people out there making special lines and dots with the various fluids in their cars.
Click to expand...

Let me be more perfectly clear then: it DOES matter. Simply throwing some thermal compound down with no rhyme or reason to its application doesn't get adequate results. Air bubbles, too little, too much can have major affects on temperatures.

FWIW, a cars cooling is nothing like a PC's cooling. Comparing the two is retarded, at best.
 
I never thought it mattered...
Until I put just a smidgeon too much on my latest C2Q build. Result = zero overclock
Took some off, did nothing else = 3600Mhz

I think if anything it matters more now than ever.... (45 nm speaking)
 
toddw said:
I never thought it mattered...
Until I put just a smidgeon too much on my latest C2Q build. Result = zero overclock
Took some off, did nothing else = 3600Mhz

I think if anything it matters more now than ever.... (45 nm speaking)
Click to expand...

Going from 0 to 3.6GHz overclock, and I can gaurantee you there was more going on then just a smidgeon of too much compound. It is FAR more likely that your HSF was not securly mounted the first time around.

I agree with retrorocketz personally. Just dab some on there and put the HSF on, done. Of course you don't want to put too much or too little, but to concern yourself with technique is pretty trivial.
 
HydroBudz said:
Let me be more perfectly clear then: it DOES matter. Simply throwing some thermal compound down with no rhyme or reason to its application doesn't get adequate results. Air bubbles, too little, too much can have major affects on temperatures.

FWIW, a cars cooling is nothing like a PC's cooling. Comparing the two is retarded, at best.
Click to expand...

yeah, it matters SO MUCH that intel sells processors with concave and convex heatspreaders. :rolleyes:

not that it doesn't matter AT ALL, because sure, you can goop it up and fux yourself over, but it doesnt matter to the extent that people make it out to.

retarded at best huh? More like you had no answer that could dispute the logic involved in the assertion.

I want FACTS from the people who say it does matter. Actual experiments with indisputable proof that shows that putting it in a line is better than just touching it on there and rubbing it around a bit. Lets see if putting on the same amount in different ways creates some magical difference.
 
I used a potato for a HS. I cut it in half then lapped the wet side with 220 grit sand paper slapped it on the CPU and turned that bitch on. It ran real cool for about a minute then slowly started getting hotter until it turned off. I think I used too much AC5 and thats why it didn't work. I think an apple might be a better HS.
 
I have built quite a few rigs and modified numerous video cards with various heatsinks,etc. and I used to use AS5.

I now use only Ceramique manufactured by Artic Silver. The best thing about this product is that it is non-conductive and independent testing has proved it to actually provide better thermal & cooling properties then AS5.

http://www.arcticsilver.com/ceramique.htm

Just my 2 cents !

Regards
 
I put down a pea sized dot in the middle... then use the edge of an index card to spread it to the edges of the cpu... makes a really thin 100% covering coat.
 
The whole theory of thermal grease is for the grease to fill in the imperfections in the mating surface between the CPU and the Heatsink with something that conducts heat better than air.

It is impossible to put "too much" grease on as it would simply ooze out the side and make a mess. It could not possibly imped the flow of heat as what it does is conduct heat. If you got some debris on the mating surface, like a hair or a fingernail clipping, then that could case a dramatic impediment to heat flow.

I suspect that that's where most of the "Too Much Grease" myths come from. They get some kind of contaminate in the grease, get some high temps, scrape some off or reinstall with less grease, and presto, everything works great. "By giminy, I must of had too much grease on that durned CPU.:rolleyes:

Don
 
Azureth said:
I have Arctic Silver 5 and am going to use it on my Intel Xeon E3110 with my ARCTIC COOLING Freezer 7 Pro HS. I've watched some videos online but they've all said different things, some say to apply thermal paste to both your processor and your HS, some say to apply it and spread it around, while some say to just put a dab in the center of your processor and leave it at that.

So my question is, what is the proper way to do it? Does what kind of paste and HS, or processor I use matter?
Click to expand...
Being someone who also owns a e3110, let me help you out.

Do not use the 'dot' method. You simply won't cover the entire heat spreader when you press down the heat sink. I did that, snapped it down and pulled it off, only about half gets covered.

Don't spread it around. A lot of people do this but you need to remember, unless you have an extremely steady hand and a very strait edge, your going to not get it even and when it's not even that means there are high/low spots, some spots might be held away from the cpu. I've tried the spread method in the past, snapped down the heat sink then pulled it off later to see only a portion of the paste even was touched by the heat sink.

What I recommend is the 'line' method. You'll see a little inlet/notch in the heat spreader on your cpu. Draw a line of thermal gel up from that to the other side of the chip. Drop the chip in, put your heat sink on, twist is a few times and press down firmly and while holding it down, simultaneously snap 2 corner of the heat sink down, then the same for the other 2 corners. This method insures there is contact directly over the cores which go up/down like that, and the rest of the paste gets pushed away, and put where it needs. The excess comes out the side. This way it's exactly where it needs to be and not where it does not need to be.
 
retrorocketz said:
yeah, it matters SO MUCH that intel sells processors with concave and convex heatspreaders. :rolleyes:

not that it doesn't matter AT ALL, because sure, you can goop it up and fux yourself over, but it doesnt matter to the extent that people make it out to.

retarded at best huh? More like you had no answer that could dispute the logic involved in the assertion.

I want FACTS from the people who say it does matter. Actual experiments with indisputable proof that shows that putting it in a line is better than just touching it on there and rubbing it around a bit. Lets see if putting on the same amount in different ways creates some magical difference.
Click to expand...

If you create air pockets between the HSF and CPU from not having an even spread you most certainly will see a difference in temperature. So if you, as you stated, just rubbed it around, in turn creating high and low spots, then just slapped the HSF on there, there is a good chance you'll have air pockets. FYI, dead air space is an "insulator". Also, applying too much and having it excessively ooze out the sides, essentially, insulates the sides of the IHS since the compound is not transfering heat to anything.

So, yes, technique and proper application can have significant effects.

Oh, FWIW, there is no logic in your comparison between a car and a PC's cooling. Even with water cooling, to consider a comparison you would need to surround the processor with coolant since coolant in a vehicle circulates through the engine block around the cylinders. There are no waterblocks or points of thermal interface material in a vehicle.
 
I spread it out and was idleing at almost 50. Same with the dot. Used the line and it significantly dropped the temps because the paste covered it all equally.

I pulled my e6300 and e3110 off and redid them each a few times to test different methods. The line method was the most reliable.
 
So basically all I need to do is spread the paste vertically, put the ACF7 on top and screw it in? Is there any clips or anything else I need?
 
An unlapped heat spreader on the CPU is pretty uneven. When you put a line of grease down the middle in the direction they tell you, you will make sure the gaps over the silicon dies on the chip have the enough grease available to get completely filled. Also, as it spreads from the middle to the edges, it will reduce the chance of getting an air pocket trapped under the HSF.

The directions I am referring to were in an earlier post in this thread.

Don
 
I'm getting ready to throw my first build together and am using the AC pro as well with the E8400. I've read numerous ways as well. The AC pro has MX-2 on the bottom of it already and thought I read that since that is already on don't worry about the chip?

I bought a tube of the MX-2, so should I clean the AC pro off and put a line on the chip only or throw a spot on the chip and heatsink?

p.s. mx - 2 says to put a spot on the middle of the chip and rotate cooler?

:confused::confused:
 
_CJ_ said:
I'm getting ready to throw my first build together and am using the AC pro as well with the E8400. I've read numerous ways as well. The AC pro has MX-2 on the bottom of it already and thought I read that since that is already on don't worry about the chip?

I bought a tube of the MX-2, so should I clean the AC pro off and put a line on the chip only or throw a spot on the chip and heatsink?

p.s. mx - 2 says to put a spot on the middle of the chip and rotate cooler?

:confused::confused:
Click to expand...

i just left the paste that came with the ac7. my temps are 25-30idle and never above 50.
 
What I've found is the line/drop/spread methods don't really matter, but what DOES matter the most, is the mount.
 
What? I didn't realize AC7 came with paste. Hmm....Would it still hurt to put paste on the CPU?
 
Azureth said:
What? I didn't realize AC7 came with paste. Hmm....Would it still hurt to put paste on the CPU?
Click to expand...

Thats what I was wondering. I've heard it bothways, but I've heard most say just use what is on the AC7.
 
Huh? I looked under my AC7 but it looked all brown? Is that normal?
 
HydroBudz said:
Let me be more perfectly clear then: it DOES matter. Simply throwing some thermal compound down with no rhyme or reason to its application doesn't get adequate results. Air bubbles, too little, too much can have major affects on temperatures.

FWIW, a cars cooling is nothing like a PC's cooling. Comparing the two is retarded, at best.
Click to expand...

If you want to talk about half a degree (if that), then yes it matters.. otherwise it doesn't.

I've posted my test results using several processors and several different thermal pastes..

For most users, it really doesn't matter.. get a good heat sink and slap it on there with some paste..
 
The AC7 (Pro) does come with mx-2 preapplied in the base of the heatsink...

I've used the dot method in my cpu with the xigmatek and some mx-2 and results seem pretty darn good given my room's constantly fluctuating temperatures (and they don't get too low unless its in the winter or the AC's cranked up)
 
Well the bottom of mine is just copper, but there is a little smudge of it on the side, I think what happened was when I took my computer parts in to my local computer shop I included some AS5, and I guess the guy who built it scraped the bottom off to use the AS5, unfortunately it wasn't until after that that he figured out it wouldn't fit on the ABIT IP35-E.
 
I have tried both spreading and just placing a dab in the middle of the cpu heatspreader and found little to no difference. I would recommend however to only place a small amount( about 1/2 a BB) of Artic Silver 5, because placing more will not help. The spaces Artic Silver 5 is filling are microscopic in size so spreading thin as possible is proper.
 
In light of this topic, I went ahead and reapplied my AS5 just for curiosity. I had originally just placed a dot in the center, now I put a line down the middel per the instructions on AS5 website. I did this a couple days ago so it has had time to cure. Results... Identical temps, not even a 1 degree difference in my load or idle temps between the two methods.
 
RamonGTP said:
In light of this topic, I went ahead and reapplied my AS5 just for curiosity. I had originally just placed a dot in the center, now I put a line down the middel per the instructions on AS5 website. I did this a couple days ago so it has had time to cure. Results... Identical temps, not even a 1 degree difference in my load or idle temps between the two methods.
Click to expand...
I don't think anyone suggested there would be a night and day difference. Especially a dot vs. line. It's more of an issue for people who "spread" the paste out themselves. The dot method acts similarly to the line, in that both methods allow the paste to spread itself out.

In the end it depends on the shape of your CPU's heatspreader and your heatsink base.
 
Dab and line tried both I like the dab in the center better. The line across I feel I get to much on the die.
 
You must log in or register to reply here.
Back
Top