What is the best method for spreading thermal paste and why is it the X method?

your graph literally has it tied for first?

> not showing margin of error
literally meaningless data

The die is tiny compared to the IHS, dipshit.

Attached: [Laughter Intensifies].jpg (1500x1104, 375K)

and yet the X outperformed it

Your own chart doesn't show it out performing anything. It's only a waste of TP.

it has the lowest temps, it outperformed all the other methods, this is fact, and the chart backs it up

Its idle temps are higher you dumb fuck making it obvious there is an error margin for load temps. If anything idle temps>load temps because it couldve had more load on "rice" method.

windmill of friendship

>If anything idle temps>load temps

holy shit these mental gymnastics are hilarious

Attached: 00.jpg (800x568, 100K)

>Your reading skills

Attached: 1527895741574m.jpg (1024x636, 104K)

Huge blob so every square millimeter is coated with the paste. The X method is as stupid as putting a smiley face on the IHS like what are you a kindergartener?

Lol yeah right. X is Shit. You out 9 dots 3x3.thats the mathematical way of doing it.

The differences seem to be insignificant which makes me happy. Thermal paste is probably one of the hardest parts of building a PC to mess up.

It really isn't. People just make it seem hard because every other process when building a PC is just like snapping lego's together expect the part where you apply thermal paste.

X works great because it spreads really well.
youtube.com/watch?v=qc7bCC1TmVg

X wins

Pea method is best. Only retard that disagrees is OP's faggot ass as proved by this thread.

Prove how Pee is the best when the video shows X was the only method to achieve full coverage

This picture makes me want to vomit. To much gross.

The best way is to put a tiny amount on the CPU, about as much as a grain of rice, and then use the heatsink to spread that over the entire CPU surface. Don't just squeeze the heatsink onto the CPU, glide it over the surface back and forth until the entire surface is covered.

The less thermal paste you use, the better. This method can spread it very thin.

>filename isn’t familycourt.jpg
I am disappoint.

This. Thinner is better. The paste only exists to fill imperfections in the surface.

>this is fact, and the chart backs it up
n=1

/thread

Compared to all the peafags itt who have no benchmarks to back up what they are saying

>X method
the absolute state of this board.

X is best general method, unless dealing with multi-die packages. If dealing with direct die applications, always spread to complete coverage on die. Multi-die packages, use lines that cover each die under the IHS. If dealing with direct touch heatpipes, apply paste to bottom of heatink, using a credit card to scrape away the excess in an orientation perpendicular to the heatpipes, then apply paste in X pattern to the IHS.

Attached: 1317206281839.jpg (256x192, 41K)

realistically line, X and pea are all acceptable. As long as you aren't doing the thing where you spread it around with a credit card first it literally has no measurable difference in your day to day computing

>see thread
>oh boy maybe i'll get some useful insight on applying paste, i always was so nervous about it
>bought the graphite paper to use in place of paste because brainlet
>it doesnt really matter
w-well at least the paper is reusable ;-;

You'll be glad to know the graphite paper is an overpriced meme like all the other thermal interface materials sold to the consumer PC market. That 30x30 square was sold to you at a higher price than the stock it came from. The material is made by Panasonic, and has been around for decades.

The important point is that you take enough.

Attached: paste.webm (1920x1080, 2.09M)

Ah, a fellow perfectionist, I see.

Based on your interest in achieving the perfect spread of thermal paste, accepting no inferiority, you may also be interested to know that you can improve the signal quality with any electronic device by using wires and cables with gold conductors. Gold is a better conductor than copper, so less voltage drops across it. That way the signal at the receiving end is stronger. I suggest you begin replacing all your copper with gold. Later, you can start replacing the wiring in your house with gold; this will save on your electric bill.

On to the meme pastes. Pic related is all you will ever need. A 1oz tube of GC Electronics Type Z9 paste is about the same price as the 3.5g of your favorite meme paste silver, and works just as well. I'm not surprised if you haven't heard of it. Review sites are marketers, they don't record the efficacy of all available products, so they don't delve into the industrial side. Below are numbers from a bygone era, an 80-way paste shootout, which featured this rare gem.

archive.benchmarkreviews.com/index.php?option=com_content&task=view&id=150&Itemid=99999999&limit=1&limitstart=12

Attached: 31SGoQ0p0VL.jpg (400x211, 12K)

>80-way paste shootout
If that link goes anywhere other than to a video of your mom getting railed by 80 dudes, I'm going to be severely disappointed in you, user

I'm sure they just wiped it off after this but its quite amusing

>overpriced meme
>paper is $15 for 30x30mm, fits standard processors
>can be reused upwards to 60 times according to my buddy
>most pastes ive seen are roughly 5 bucks
>can only be used once and according to another friend must be changed out infrequently
>performance is the same as paste as far as i've seen
wut

Sorry to disappoint. Back in the day tech product "shoot outs" use to be a common thing.

Thermal paste is an atrocious conduct, barely getting to 1 digit K⋅m⋅W−1. Its main function is to displace air, an even worse conductor, so it doesn't form an isolating cushion between the IHS and the heatsink.
Pic related; the average piece of paper is around 100um thick for reference. The X method uses a stupid lot of paste that will not get displaced unless you put some serious pressure over it.

>sub 1 degree differences
wew, at that point you're practically inside the margin of error of the average sensor. Bumping your heatsink's fan an extra 100 rpm would make much more of a difference.

Attached: Impact-of-thermal-interface-material-TIM-thickness-to-silicon-die-temperature.png (850x581, 91K)

see
$13 shipped on Amazon and you can easily get more than 60 applications out of a tube. I'm sorry you're dumb and allow yourself to get jewed.

Who cares about CPU/high pressure heatsinks methods? X works best but the difference isn't that much either way
What matters it's low pressure and bare dye comparisons, there barely any out there
ZP-360, aka 0.99 USD paste, on eBay and AliExpress lasts dozens of applications and performs equally well if not better
Enjoy you air bubbles
Keep buying overpriced and worse performing thermal paste dumb goyim
Huge blob only works fine if the paste has low viscosity and the heatsink mounting high pressure
X still works better than pea on CPUs, most heatsinks exert enough pressure to displace most pastes

>5um
>max is less than average

Came from a paper by the IEEE, so I guess that they have the ability to easily determine the thickness of the paste, they're no amateurs or techtubers. But you're right about the average, I accidentally let out a piece of text that referenced the image and said "average temperature," not "max temperature." I guess the image has the typo.

Why not get yourself some low melting point metallic alloy and solder the ihs to the heatsink if you're not planning to change your cpu in the near future?

god damn that benchmark is old

Swastika

>What is margin of error HURR DURR

Attached: 1517811358757-tv.gif (495x359, 1011K)

man just use whatever fucking works, i'm not sure why people have to overcomplicate this. all of them get the job done just fine, i use the pea-ish method and just dab a lil on there. i swear people find a way to argue over anything

...and this makes the data invalid how?

Attached: 1517631959883.png (640x640, 36K)

it doesn't invalidate the benchmark, but there are definitely better products on the market nowadays compared to 2009

most of the fillers used in modern pastes are the same as in that benchmark
there's been no huge breakthrough in materials technology that fit this extremely niche application

>being this new
Nope they don't

differs with CPU brah seriously

Attached: gn-modmat-gallery-05.jpg (1000x667, 201K)

>2 dollar difference is getting jewed
>nitpicking this hard
at that point its really a matter of personal preference, im sorry your fee fees got hurt because your speshul paste™ isnt the objective best for everyone

Attached: you_that_you_so_desperately_crave.gif (480x238, 415K)

>he's still using deprecated pleb paste

Attached: 81zhua2retL._SL1387_.jpg (1000x1387, 293K)

It will take a decade for you to get your investment back on this thing, while performing a few degrees worse

Come on, Jow Forums. Post that gif, I know you want to.

where is the unsatirical msi tuto of applying thermal goo, time has come to save that thing for later use

Attached: 1525137847856.webm (800x450, 2.11M)

Why is the heat spreader necessary? Why not a perma delid?

>Tech fag from gayernexus.
I love how g tell ppl to go to v if a gayme is mentioned, but you are promoting a channel that targets gaymers.

apply it according to msi tutorial

t method > x method

Fact

what is the best method for OP to commit suicide and why is it the vertical line method

isn't the whole point to cover as much as possible without spilling? just use a qtip and spread it (thin close to the edges)

>literally 2 degrees
This thread is worthless and you are a worthless human being

>As long as you aren't doing the thing where you spread it around with a credit card
/thread

what's wrong with credit card method?

People say it causes air bubbles but in reality the pressure exerted by modern CPU coolers will force them out if there is any. Though if you want to be super safe other methods are probably better.

you started with less than a pea.... you had a sweet baby pea, plus ye ihs is larger than die

you started with less than a pea.... you had a sweet baby pea, plus ye ihs is larger than die by alot. i.e. go watch 8700k delid its like a third of the overall area

this is the utmost dumbest shit i've seen, with the right amount of coverage and lack of air bubbles the will all perform the same due to pre-spaced coolers only allowing a set amount of squish.... the question is did you get enough coverage without leakin thermal nut onto your cpu socket

that ryzen is wrong, the die underneath is oriented the same way as the text is

Just another reason X is the best and most reliable method

>worrying about muh investment when ease of use is worth it

That is the point of that thing is that you can place it and never have to worry about it for the lifetime of the PC. It'll be a few degrees worse but it will be a few degrees worse forever unlike paste which will dry up and be way worse than the pad ever could be.

Replacing paste every 2-3 years over the IHS is not difficult and only takes like 30 minutes tops, I'd rather have the performance than the laziness and poor performance of the pad

The pad is only 1-2 degrees off. Sure if your an extreme overclocker it's junk but for nearly everyone else it's a nice fire and forget solution and it may even be better than paste if you perhaps did a bad paste job last time it was done.
If your hurting for that extra 2 degrees then the pad is bad but for basically everyone else it's overall better

>Huge blob only works fine if the paste has low viscosity and the heatsink mounting high pressure
You might be right. The thermal pastes that I use aren't that viscous.

The video benchmark I saw showed it off by 4C, sometimes even 6C, thanks I'll stick to pastes with that performance

Paste or liquid metal?

Do not use liquid metal if you don't know what you're doing.

use a credit card or something like it

>Not using the superior smiley face technique

Attached: pic_disp.jpg (800x600, 77K)

If you have to ask, use paste.

Arctic Silver 5 was king of the hill 10 years ago

Not so today

Has anyone told you that you're a kill joy because you sound like a kill joy.

AS5 is pretty garbage compared to kryonaut, nth1, or maker nano

>Athlon 64
>Abit motherboard
Even when they made this a few years ago nobody cared about any of these components for more than a second

Mounting a fuckhuge cooler directly on the die is risky business, IIRC IHSes only became a thing in the early 2000s around pentium 4 era

*Cracks your die*

Attached: 7PJ1Sq3.jpg (1804x608, 285K)