Thermal Compound Application Methods Benchmark

[nicklmg]

Amazon: http://geni.us/1kJp

NCIX: http://bit.ly/20WqgrI

 

Will your thermal paste application method impact your CPU temperatures?

 

 

 

[zMeul]

I was right, line method is better

[noisebomb44]

7 minutes ago, zMeul said:

I was right, line method is better

straight from the underground

[dragoon20005]

47 degrees all the method except too little

 

kinda expected

[QueenDemetria]

I can hear it now: "But Luke, you didn't use a static fan/pump speed, which can alter the results. If the fan is adaptable and has a target temperature it will change the fan speed to achieve 47c. If the fan/pump speed is static it removes more variables."

[Progressor]

Come on Luke, how can you test this at 47 degrees C and expect any results.. Sometimes it looks as you guys have no idea what you're doing. The purpose of a thermal compound is to transfer heat. HEAT, yes guys HEAT! You need to push the thermal paste to it's maximum heat transfer limit, that's when other factors such as spreading start to make the difference, at 47 C there is simply not enough heat transfer for the spreading method to "bottleneck" the thermal paste. I'll repeat that again, you need high enough temperatures(80 C should do) so the spreading method becomes a bottleneck(at least the incorrect applying methods) to the heat transfer! Yes, all of this probably wouldn't matter for someone who has a CPU with good TDP or runs it at stock but for an overclocker or someone with a bad cpu it would make enough difference(GPU usage applies too). As someone who has to deal with an old overheating laptop(which if I didn't mod correctly wouldn't even boot into windows before it overheats), everything makes a difference when it comes to high temperatures.

[EnduringGuerila]

You should have tested with a less efficient CPU cooler. Like the 212 air cooler. Or even test with a stock cooler. Just use anything that might be more affected by the paste. It is fun watching Luke lose enthusiasm as he keeps going though..

(I posted this in the video comments but figured it would be easier to discuss here..)

[bobhays]

i liked the music on this one unlike the last video

[JoshB2084]

34 minutes ago, bobhays said:

i liked the music on this one unlike the last video

That's explain your Avatar!

 

For thermal paste, I using the dot method.

[NickTheMajin]

Haven't watched the video. 

 

Answer is no it does not matter unless you dump the whole fucking tube on the chip or don't use it at all. 

[gonvres]

So far from these videos I'm learning

 

- How many case fans you have doesn't really matter all that much. 

 

- How you manage your cables doesn't really matter very much.

 

- How you apply thermal paste doesn't really matter very much.

 

Its like debunking all the trends computer enthusiasts obsess about.

 

The big one I'd like to see; Is water cooling actually quieter than air cooling? As we know that most fairly small AIO water coolers don't produce any better temperatures than mid end air coolers (212 evo or better), they always run the risk of leaking all over your system, it seems the final line is "well, they're quieter" but I don't think thats really true either considering they often have cheap fans and pump noise. It seems to me you're nearly always better off saving money and getting a 212 Evo or for equivalent price to a single rad AIO CPU cooler getting a tower noctua cooler.

[gonvres]

20 minutes ago, NickTheMajin said:

Answer is no it does not matter unless you dump the whole fucking tube on the chip or don't use it at all. 

True story, when I had a system built by a small computer shop before I felt comfortable putting it together, they put 0 thermal paste on it, the cpu ran without any issues for multiple years in this state. 

 

Not that i'd recommend you do this at home.

[LucidMew]

On the results page, it says it's a 5930K, but yet, the CPU that Luke was using looks like a 6700K/6600K? 

[xnamkcor]

You forgot "Licking It", "Molten Copper", and "Naked".

[maatriks]

So that was a bit anticlimatic. All those years I have spread the paste evenly on the CPU. Oh well

[nobelharvards]

I think the best thing to take away from this is... Companies are not completely blind or stupid. They know more people are getting into building their own rigs and dabbling into the entry - intermediate level of complexity, therefore they have put a bit of money and research into idiot proofing, such as making sure cooler mounts spread the paste to near equal levels in most mounting attempts.

 

Like others have said, perhaps there would be a difference if lower quality components are used, so the thermal interface is the bottleneck. Or just use older stuff that may be lying around. Older components will be less idiot proof in general. (E.g. When Intel first added temp safety mechanisms and AMD did not add it until later.) 

 

 

Video is an old video from tom's hardware or something,  from 2005/6.

[aflyinghotpocket]

I would like to see a followup to this video Where Luke tests the following:

 

• Different types of thermal paste (are some better then others?)
• Different types of cooling methods (liquid vs. fans)
• Completely naked CPU (no thermal paste) with and without cooling 
• Different CPU?, AMD chips tend to run a bit hotter then Intel (in some situations). 

I know there are WAY to many combinations to cover in a ~10-12 minute video, but I would like to see more variety with this Workshop. 

Great video as always guys!

[EricX2]

It's almost like people don't understand what happens when you apply pressure to the heatsink. Spoiler, it spreads the thermal compound out as far as it can based on the amount not the shape. As long as you put the right amount and it's in the middle so it can spread the results will be the same.

[SolarNova]

good vid ..although..

 

I would have tried at a higher temperature. get to OC'ing, higher temps put more emphasis on the TIM, and consequently its application method.

 

Also, I didn't notice but did u ensure static fan speed ?

 

And finally, maybe 3 different TIM's should have been used, a liquid metal TIM, a runny paste, and a thick paste, they will all behave differently and the application method will likely impact them differently aswell.

 

I personally use the spread method, iv used the 'card method' but with a proper spreader, and also the rubber/plastic gloves (or bag over finger) method. I find it more reassuring knowing there is a even layer of TIM over the entire IHS. My next build, of which im gunna be making a proper build log for as its gunna be rather high end ..many thousands of £££, will be using liquid metal TIM and ill be spreading that also.

[brandishwar]

3 hours ago, nobelharvards said:

Like others have said, perhaps there would be a difference if lower quality components are used, so the thermal interface is the bottleneck. Or just use older stuff that may be lying around. Older components will be less idiot proof in general. (E.g. When Intel first added temp safety mechanisms and AMD did not add it until later.) 

Yeah that was my thought as well with some of the other workshop videos they've done. They are doing it with all modern parts and components. Which is fine given that's what's on the shelf, so that is what builders will use. But as I said in the thread on the cable management video, things are much easier today when it comes to building a system. Cable management doesn't matter nearly as much because of 120mm fans, and the thermal paste application doesn't matter much because of adaptations to socket and processor designs and the inclusion of various safety measures to keep the CPU from cooking itself.

 

The video you linked shows not only why thermal paste application was so important 10+ years ago, but it also shows how AMD and Intel along with the motherboard makers responded to published reports of people literally frying their CPUs doing stupid things. I remember an article in PC Gamer magazine, this was 1999 or 2000 I think, where someone took pictures of a socketed Pentium III system they built and forgot to plug in the CPU fan. The system took not even 10 seconds to literally fry itself. The mainboard, processor, and memory were all gone. 1000 USD in hardware destroyed by one oversight.

 

You can't really do that anymore. The BIOS will throttle the multipliers into the single digits and the system will power off if the processor gets too hot. There's another build video of a guy building a system with a 3960X and not putting the cooling solution onto the processor before turning it on for a POST test. About 45 seconds later, the system powered off. That was long enough to show everything working, and the BIOS kept the processor from frying itself when it exhausted all mechanisms for keeping the processor from overheating.

 

And the video preview image above also shows the rubber standoffs that AMD included on their processors that Intel -- at least initially -- did not. And that was in response to reports of people crushing their processor cores trying to mount the cooling solution. There was a special process you had to take to mount it, and you had to make sure not to press down too hard trying to clip the processor onto the socket. When I ordered an Athlon XP board and processor, the company from whom I ordered it had a service where, for an additional 20 or 25 USD, they would install the processor and cooling solution onto the board before sending it to you. Knowing what I'd read about people crushing their processors, I willingly paid the extra. At least if the installer crushed the processor, it was entirely on them and they swallowed the loss, not me.

 

The eventual better response was the integrated heat spreader, which Intel switched to before AMD. It kept you from crushing your processor and made the application of thermal paste not nearly as important. Though Intel would, for a while, get stupid and put ordinary thermal paste between the IHS and CPU die on their processors instead of using solder to ensure maximum transfer of heat. I hear with Haswell-E and Skylake they're using solder. At least AMD has always used solder, to the best of my knowledge.

 

The other big problem with this test was using a more power-efficient processor -- Skylake processor is what that looked like -- that just isn't going to get nearly as hot under load. In actuality, this test should've been conducted with an AMD FX processor -- bonus: make it an FX-9590. Given how hot the FX processors can get under load, that would've shown quite easily how much applying thermal paste can matter in the right circumstances. Or, better yet, go with a Pentium 4 that had the pure copper heatsink that had to be bolted through the mainboard into the tray behind it.

 

The better conclusion on the video isn't that thermal paste application method doesn't matter, but that it no longer matters to the degree it once did. Instead, and this is as true today as it always has been, quantity and quality of thermal paste is what matters.

[Manage My Cables]

As long as you have a even coating across the heat spreader, you should have any problems. I personally use Thermal Grizzly and that comes with a screw on applicator. 

[jasonvp]

3 hours ago, aflyinghotpocket said:

Different types of thermal paste (are some better then others?)

I'd pay money to see this one.  TIM manufacturers don't have as much at stake as far as how you apply the product.  But they certainly do when it comes to which product you choose.  I'll bet the results from the testing would ruffle some feathers.  My gut tells me that it doesn't matter much. ;-)

[Razor512]

From repairing many laptops, I found the spread method to be most effective in cases where you have a bare die, as with other methods, you either get an uneven spread where some portions of the die will have no compound (stability issues when temperatures seem fine), or you get unnecessary spillover in order to get good coverage.

 

For CPUs with a heat spreader, good cooling can be had as long as the die area is properly covered.

 

The CPU die can come in a wide range of shapes and sizes.

 

 

And some more ancient artifacts:

 

 

 

[Voigt]

This was one of the videos I had sadly to dislike.

 

Like many other already said there were alot of problems in this video one beeing: You used a "to good" cooling, and to little power usage on the CPU.

With a AMD FX9590 (The 5Ghz Stock Model) with 220W TDP you would get different results then with a modern Mainstream Intel one.

And I know you probably don't have this model, and won'T buy a 800USD CPU just for a workshop episode, but a modern AMD FX CPU, oced to the max would also be sufficent.

Although the conclusion would be mostly the same, since the tempurate would only be differing about 1~2K nothing major.

But with your "test" this concluded to be a very boring video.

 

As a general nice to know, and also as info for the people who ask about different thermal paste: 

Although this is a German Video, you still get the jist, since you see which CPU they use, what "thermal compounds" and what the results are in the graph later on.

[DXMember]

In the intro you promised X shape, there's none in the video

Where's the circle?

Where's the smiley face?

 

Learn from the best?

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

 

also where is CISCO application method?