best cpu thermal paste

[pegasus1]

Really?

So people who want decent temps ignore the industry advice of using a blob or a line and spread the paste all over the die

People who want decent temps generally listen to those that know and as far as im aware there are no industry reps present during extreme OCing events, just enthusiasts.
Read further up the post, people like FatRakoon know their stuff.
For me and my lapped CPU's/HS's its a translucent layer on both surfaces.

Was a chilly morning in Essex i can tell you.

 

[FatRakoon]

Im not an OCer in the Household Division

Sorry, I didnt think.

When you bull your boots to get them to shine like a mirror, you use small circular motions ( roughly 5p to 10p sized circles ) and you do this to your army boots with your kiwi polish and this is called "bulling".

Well, in that same way Id bull my boots is the same way I would apply heatsink grease.... It ensures an even layer across the whole surface, or as even as I want it to be anyway.

 

[pegasus1]

Sorry, I didnt think.

When you bull your boots to get them to shine like a mirror, you use small circular motions ( roughly 5p to 10p sized circles ) and you do this to your army boots with your kiwi polish and this is called "bulling".

I meant that most of us are either wearing dirty boots on exercise getting ready to deploy, wearing desert boots whilst deployed or wearing smelly trainers on a drunken Monday lunchtime once we're back.
But i get the bulling a CPU thing, its what i do with my TIM rather than my boots, not that kind of soldier

 

[FatRakoon]

its what i do with my TIM rather than my boots, not that kind of soldier

1 - Its what you do with your TIM?

2 - You are not THAT kind of soldier?

I think that if thats what you do to your Tim, then I think that you and tim, are both very much THAT kind of soldier!!!

 

[pegasus1]

1 - Its what you do with your TIM?

2 - You are not THAT kind of soldier?

I think that if thats what you do to your Tim, then I think that you and tim, are both very much THAT kind of soldier!!!

Ive been outed on the interweb, aaaagh.

 

[FatRakoon]

ROFL

 

[Finguz]

People who want decent temps generally listen to those that know and as far as im aware there are no industry reps present during extreme OCing events, just enthusiasts.
Read further up the post, people like FatRakoon know their stuff.
For me and my lapped CPU's/HS's its a translucent layer on both surfaces.

I know my stuff too, and piddling about spreading TIM is not going to help anyone

Anybody telling you that spreading a translucent layer of TIM is going to be more efficient than squirting a blob of the stuff does NOT know what they are talking about

P.S Im not sure where you have been going but every overclocking event I have attended has been crawling with Reps.

 

[FatRakoon]

I disagree.

Simply putting a blob onto the CPU will end up insulating the CPU from the heatsink...

The whole idea of TIM / Gloop / Colgate / Jizz / Whatever you want to call it is simply to "Fill in the cracks and gaps left behind, and ensure that the contact between the CPU and the Heatsink is an absolute minimum... Plonking a blob of compound in the middle of the CPU sure enough does do its job, as expected, but spreading a thin layer will ensure that the Heatsink will be as close as is possible to the CPU in order that it can do a more efficient job of cooling the CPU.

Now, dont get me wrong here, I firmly believe that everyone has their own way or doing things, and I do believe that they can both correct... ( Layer or blob ) but as long as there is some there and as long as it covers most of the CPU, then whats the problem?

Although I do think that a "Translucent" layer is too little, I think this was a simple joke meaning that its spread thin.

I would suggest tests. As anyone who knows will know... The tests cannot possibly be done instantly, so, if anyone prefers to blob their CPU and they know their temps right now ( IDLE / LOADED ) then could they try a thin layer on the whole of the CPU and HEATSINK, and after a week, let us know the temps? - indeed, for those who like to spread it properly, if they could just slap a blob on and again, in a weeks time, let us know the results there?

I know for myself, I spread it and I have done thsi purely because when I used to plonk the stuff on, it was fine and it did the job, but I needed to replace 3 coolers and I had near to no gloop left and so I went a little stindgy and make it spread and I saw the temps were good, btu after a few days they had settled in and in all 3 PCs the temps had dropped by around 2-3c, just by spreading the stuff thinner than I had been doing.

I suggest that we do some tests before mouthing off at people and tellign them they are wrong when there is far more evicende to prove that a thin layer is better than a thick layer.

 

[Hesky82]

If anybody is telling you that spreading a translucent layer of TIM is going to be more efficient than squirting a blob of the stuff does NOT know what they are talking about

I agree, the average retention mechanism for a cpu heatsink designed to support a processor will create a static preload pressure of between 40 lbf and 50 lbf. That’s a ***** load of pressure! Add to the fact almost all these different thermal compounds we use today are low viscosity greases we will find that the thermal compound is forcefully spread/squeezed out to an identical layer thickness regardless of how it was applied to the cpu in the 1st place. I don’t care if you use a blob, line, circle, spread or whatever, providing you end up with enough thermal compound to cover the ihs/chip and create no air pockets in the process you will end up with the same thermal interface performance.

As far as I’m concerned, heatsink clamping force affects thermal compound performance significantly more than the method used to apply the thermal compound.

 

[Finguz]

I disagree.

Simply putting a blob onto the CPU will end up insulating the CPU from the heatsink...

I disagree.

I have been building PC's for many, many years and have always used a blob, splodge or lump of TIM and all the PC's are running cool and quiet. The pressure between HSF and CPU surface spreads the TIM far better than you and some kind of 'mini spatula' will ever manage.

Another point you would do well to understand is that the TIM cannot be spread and worked into all the surface imperfections, cracks e.t.c @ room temp, it needs to be at operating temperature and in some cases it needs time at operating temps to 'cure' before it will work properly/at it's best.

How you spread it when it's cold has no bearing on any of this

 

[FatRakoon]

Not all coolers apply as much pressure as you think.

Although I fully understand that the pressure in most coolers is more than enough to spread the stuff, providing enough gloop is used, and also you seem to miss out the fact that some gloop is thick as pig No Swearing! and this stuff does not spread under the force of the heatsink very much at all. Crappy white stuff all spreads fine sure enough, but many times I have seen CPUs with a blob of gunk thats been wedged under the heatsink and its almost half a mill thick and not at all spread over the CPU! - Sure on the older CPUs like the XP / TBirds etc, this would pobably be ok as it would cover the entire surface, but a 10cm half mill thick pile of stuff is simply not good enough... Again, thinner gloop and it would have probably been ok.

- I dont care how long you have been building PCs. I run my own business doing it and have done so for 4 years and I have been building them since my first PC in 1986/87 ( Amstrad PC1512 2xDD and later upped it to 20MBHD/3.5+5.25, 512K to 640K, CGA to VGA etc ) and I have been repairing ( 8 and 16 bit ) computers since before that.

- I dont use anything other than my fingers to apply the grease


Listen, this is really getting out of hand.

Its also getting ridiculously immature in that people are starting to **** each other off over the way they apply heatsink compound?? - how bloody retarded do we have to be, to argue over that???

You have your preference, I have mine, the fact is that they are both correct does not seem to click with anyone here.
Sure enough, the blob method is fine and yes, I suppose when I have a number of systems that Im knocking up, then I do just plonk a blob, and sometimes if the system is only a stock basic system, the compound on the heatsink is more than adequete, but when I do a high end system or a clocker for example, I just feel more comfortable knowing that the whole of the CPU and Heatsink are both coated with a good layer of good quality stuff.

 

[kmufc77]

i started this thread a while back and wow what a response i used AS5 and im not getting the temps i was hoping for. i have a swiftech H20 Appex 220 kit cooling both my GTX and my Quad 6600 Be revision. my GTX wont go over 45c using the dtek waterblock.However my quad idles at 47-50 i have applied the paste using the artic silver instrutions on the website putting a small line across the cpu which somebody else on here shown a diagram. im guessing either the swiftech cpu water block is at fault the rad isnt up to stratch cooling both my GTX and 6600 or the paste is applied wrong but thanks all for the great response.

 

[Finguz]

- I dont use anything other than my fingers to apply the grease

Well that's a particularly bad idea as your finger tips are very greasy and will contaminate the TIM.

Thick stuff that doesn't spread when it's cold should give you a big clue as to why it doesnt matter what the TIM is like at room temp because all the action occurs as the temps increase and the TIM becomes more fluid and the pressure between CPU and HSF will work it into all available areas.

This is not getting out of hand at all, some bad advice has been given and I feel it's important to refute it.

 

[Tuvoc]

Maybe he uses his fingers with a bit of shrink wrap over them - but point taken

What I do is use an old credit card to put a translucent ultra thin layer across the top, then a very small blob in the middle. Hopefully best of both worlds...

 

[AgedIdiot]

Although off the original ? , which was 'best cpu thermal paste' the comments about how to apply I found interesting as I've never been able to work out if there is , in fact, an absolutely correct approach (with all others being wrong!).
As it was such a long time since I'd looked it up I had a look at Arctic Silvers web site fot their advise/opinion and I've provided a link below

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

If I interpret correctly they go for the blob approach on all Amd cpu's BUT for Intel Dual core a bit different and for Intel Quad different again.

P.S Do remember that at one time it was recomended that you 'spread' not only the cpu but also the cooler base. - May have been a cooler manufacturer or two cannot recolect now

 

[Steves_pc]

AS5 here too.
infact i use it on every build, fwiw i use the end of the tube to dab it out over about the central 10mm square then give the hsf a wriggle about before clamping down then leave heat/clamping force to do the rest, when i've rmoved the cpu's i've always found the compound to be out to the edges just nicely.

I too have been building/repairing systems since way back in a time where oscilloscopes/soldering stations were used more than just screwdrivers and i'd say no method is wrong just so long as you end up with a sufficiently thin coating over the contact area one way or another, as several above have said too much only makes matters worse.

@Finguz - on the occasion i get as5 on my figers it sticks to me like crazy and not the hsf.

 

[monkeyspank]

I rub some into the base of the heatsink and then put a properly sized blob in the middle. And then beast it with some heat

 

[Renegade87]

I usually spead it myself nice and thin, attach the heatsink correct, remove and see if more is needed on the areas where there was no contact (assuming it was fitted correctly in the first palce, I now get 47oC Load with my 8400 at 4ghz so it must be working.

 

[Finguz]

AS5 here too.
infact i use it on every build, fwiw i use the end of the tube to dab it out over about the central 10mm square then give the hsf a wriggle about before clamping down then leave heat/clamping force to do the rest, when i've rmoved the cpu's i've always found the compound to be out to the edges just nicely.

I too have been building/repairing systems since way back in a time where oscilloscopes/soldering stations were used more than just screwdrivers and i'd say no method is wrong just so long as you end up with a sufficiently thin coating over the contact area one way or another, as several above have said too much only makes matters worse.

@Finguz - on the occasion i get as5 on my figers it sticks to me like crazy and not the hsf.

Hehe yeah that's the grease on your fingers

To an extent I agree that no method is particularly wrong ( except using your fingers which is just silly - you may as well spread it with a lard covered sausage ) as long as sufficient compound is applied.