Reminder to never buy Asus GPUs. Ever.

>>55687213
why?

>>55687213
nice try competitor brand shill

>>55687213
i'm not sure what this picture/op is trying to convey

>>55687228
Do you not see that only 3 out of 5 heatpipes would be in contact with the die?

>>55687244
Sure senpai, I'm just trying to warn you...

>>55687251
Are you blind?

>>55687257
yes

i have a brail keyboard and i stick a displayport directly into my forehead

File: cooler2.jpg (674KB, 900x1198px) Image search: [Google]

674KB, 900x1198px

>>55687264
>dat thermal paste around the die

Not the worst I've seen lately. Pic related is EVGA GTX 1070 SC.

File: DSC_8986.jpg (466KB, 1200x798px) Image search: [Google]

466KB, 1200x798px

THEY DID IT AGAIN

Who gives a shit as long as it gets good temps

>>55687257
Yeah, but explain why that matters? As long as it doesn't overheat who gives a shit?

saged

File: a26x79.jpg (63KB, 800x600px) Image search: [Google]

63KB, 800x600px

>>55687257
>Do you not see that only 3 out of 5 heatpipes would be in contact with the die?

Yea, but pipers are touching each other, acting like heat spreader on CPU.

>>55687325
>>55687326
>>55687332
This. OP was a fag once again.

>>55687213
>>55687315
if the pipes are in contact the heat will transfer to all the pipes anyway even if it's not directly from the chip.

would you rather have too little thermal paste of too much.


?

>>55687257
>only 3 out of 5 heatpipes would be in contact with the die
>GPU die still has full contact with the heatpipes
>heatpipes can still pull heat away from the die
>heat can transfer from one heatpipe into another if it really needs to
>Cooler still keeps the GPU cool
So literally nothing then.

File: gay tv lol plasma.png (101KB, 784x691px) Image search: [Google]

101KB, 784x691px

>>55687332

If the pipes are touching its thoroughly gay.

File: asus a shit.png (611KB, 726x1246px) Image search: [Google]

611KB, 726x1246px

>>55687359
It's not nearly as efficient as having direct contact.

AND THEY KEEP DOING IT

>>55687359
They will, but not as efficiently. I thought the direct contact heat pipe meme died out years ago. It offers zero benefits over a solid base.

>>55687374
I'd rather have the right amount on my $X00 piece of electronics. How fucking hard can it be to apply some thermal paste? Even the average retard on here can do it correctly.

>>55687213
Bad Pajeet. Bad. No Rupees for you. Now go back to your hole and think of what you've done wrong.

File: asus rog matrix 290x.png (523KB, 685x576px) Image search: [Google]

523KB, 685x576px

EVEN ON PREMIUM MODELS

File: Thermal-resistance-vs-thermal-paste-thickness.jpg (14KB, 503x233px) Image search: [Google]

14KB, 503x233px

>>55687374
both are bad.

>>55687392
I'd take this, hook it up to a PC and run a program that displays a series of bright white bars at max brightness in patterns across the screen to remove it.

Nice free TV

>>55687393
Yea... But you realize that to have only direct contact they'd just remove excess heat sink thus making the whole thing less thermally conductive. This whole thread is stupid.

>>55687408
Who's he shilling for in your scenario? Not Asus Inc.™?

>>55687379
>>55687359
>>55687332
>>55687326
>>55687325
The card could be cheaper by throwing 2 heatpipes away.
It's also pretty loud considering it has that much surface area to dissipate the heat. Shitty fans.
If ANUS would be competent they would design a separate cooler for smaller die cards or add a base

>>55687428
but if you put on too much the excess will just be pushed out

>>55687448
Says who? Less pipes doesn't mean less heatsink.

>failing intro to thermodynamics

>>55687309
It literally doesn't matter
https://www.youtube.com/watch?v=r2MEAnZ3swQ
To much > to little

>>55687612
>comparing a CPU with a heatspreader to a bare die
>linking to LinusCatTips

Back to /v/, kiddo.

>>55687609
>le you don't know fisics meme

File: cooler2.jpg (245KB, 900x600px) Image search: [Google]

245KB, 900x600px

Direct heatpipe contact is a meme hyped up by manufactures to save cost.

Sapphire does it right.

VRM temps are shit though.

>>55687639
>Intel doesnt use 3 liters of cum paste + shitty glue on there dies
It's the same shit.

>>55687465
>The card could be cheaper by throwing 2 heatpipes away.

Not necessarily, the cost savings from using less material would probably be outweighed by the increased cost of design and tooling for different coolers for each card.

I'd also guess that in addition to supporting multiple models of card, they probably have some variation in layout of the same model over time, and so they design a cooler that allows for slightly different chip placement and the like.

>>55687643
>no argument

>>55687651
Is it a vapor chamber or just a copper plate?

>>55687727
plate.

Chamber is used on Vapor-X, this is TriX

File: qSzgfYN.jpg (143KB, 1200x897px) Image search: [Google]

143KB, 1200x897px

>>55687727
nope

File: 1sapphire r290 voc.jpg (58KB, 608x490px) Image search: [Google]

58KB, 608x490px

>>55687727
this is the vapor-x though

>>55687213
My xigmatek cpu cooler were like that though not a flat surface, temp is funny on core 0 and higher than core 1, the different is a lot could go like 10c+ on all 4 cpu core.

>>55687465
>The card could be cheaper by throwing 2 heatpipes away.
Did you miss the point that those pipes will still transfer heat away because of direct contact with the pipes next to them?
The cooler would literally be less efficient without them.

>irrelevant basement dwellers expose thermodynamic inefficiencies of a multi-billion dollar company's product: the post - part 2

>>55687498
... You should probably look up how heat transfer works before your post any more

>>55687213
OP confirmed retard

>>55687309
>>5568726
>>55687315

SHUT THE FUCK UP YOU BUNCH OF FUCKING RETARDS. YOU CANT HARM A GPU WITH PASTE!!!!!!!! MAYBE A CPU, CAUSE IT CAN SQUEEZE OUT AND LEAK INTO THE SOCKET AND FUCK YOUR SHIT UP, BUT IT WONT HAVE ANY NEGATIVE OR NOTICEABLE EFFECTS TO THE PERFORMANCE, LIFESPAN OR THERMALS
OF YOUR GPU SO SHUT THE FUCK UP THE ONLY ONE THAT IT NEGATIVELY EFFECTS ARE THE PEOPLE PUTTING TOO MUCH ON THERE, CAUSE SHITS EXPENSIVE BUT ANYWAY GO BACK TO FAPPING OVER PICTURES OF THE GNU/LINUX FREEKEK STALLMAN ON YOUR MEMEPAD CORE2DUO WITH MUH ARCH ON IT4

>>55687393
>It's not nearly as efficient as having direct contact.
actually it's more efficient. Heatpipes move heat better than copper alone.
...and they're soldered together, so they make good contact.

>>55687470
Assuming you mount with sufficient pressure.

>>55687213
IT'S NON CONDUCTIVE

>>55688909
cool

>>55688909
Go take your meds, autist.

File: 02b.jpg (148KB, 1000x451px) Image search: [Google]

148KB, 1000x451px

>>55687213
Plebs, get a better cooler.

'TPU review gone wrong somewhere.
The GTX 1060 in its own review 7.5% faster than the ref 480.
In the Strix 480 test it is 11% faster than the Strix, with the same games, same drivers, same FPS numbers. '

'I did some number crunching and posted it on TPU forums - he said he made an error and is going to redo the charts. '

>>55689840
Putting that chunk of metal between the heat pipes and the chip makes it transport heat less effectively

Just because you cover the pipes with a plate so you cant see how many directly contact the chip doesn't mean its better.

>>55689935
Radoen chips are placed diagonal, the plate is rotated to and fits perfectly. So all pipes touch the chip. The previous picture is probably a GTX card, probably there are some difference between models.

Also the heat spreads in every direction. Its more of a question how much copper for cooling you have than whats the contact area is. I cant calculate it (now) but i would say there isnt less heat traveling through pipes not touching the chip directly. I think the plate and pipes heat up and transport heat even.

>>55689157
>>55688088
You idiots. The point is they fucked up the overall design and just went MORE HEATPIPES when what they should have done was route them properly to fit the core.

why come they don't use diamond instead of coppy

>>55690687
desu they should vacuum weld the cpu to the heatsink

So I should go with MSI this time?

I purchased an MSI GTX 1080 Gaming X.
Did I do okay?

>>55687447
If you get into the dev menu for the tv, most have an actual burn in mode.

>>55690531
Wrong, please reread thermal conductivity.

>>55690877
Might as well have said "NO U"

>>55690939
Might as well said to fuck off but at some point your just to tired. So i pointed to the required resources for better understanding of the topic.

File: DSC_0012.jpg (3MB, 3840x2160px) Image search: [Google]

3MB, 3840x2160px

This is my 980TI STRIX.

What the fuck Asus...

>>55691015
>your just to tired

Apparently so. You're still wrong though. Proof is in the data regardless.

>>55691036
There wasn't any data posted.

>>55691080
If you're one of the posters you replied to it. Asus is the worst performing cooler, even comparing to other 2 fan coolers.

>>55689840
the one and only morpheus

>>55691107
I have no experience with asus coolers but wont deny that they provide bad cooling.
But this bad performance comes not from the fact that not all pipes have contact with the die. Heat can travel from one pipe to another. Also your contact area is fixed size which limits the total possible amount of heat that can be transported.

The bad performance comes from
-bad air flow
-amount of material used
-quality of material

>>55691348
Sure thing bud. Except for the fact that the cooler was fine when they used it on a 780 Ti and didn't have the same performance gap compared to other coolers.

>>55691415
On which gpu do they use them now ?

>>55691455
Their DirectCU II cooler has been retired but they still seem to be doing the same practice hence the OP. That Strix cooler is the exact same one they use on the 1080

>>55691471
According to guru3d the cooler performs worseat idle but better under full load. I would say they turn down the fans at idle lower then before. Also the 1080 has a lower tdp then 780ti, the cooler should perform better, which it does.

The fact that not all pipes contact the die does not hinder performance. They heat up evenly with very small differences between each other. The transport on the outer ones could even be better because heat travels better the higher the difference of temperature.

Because the tdp has fallen they might have reduced the amount of copper to lower costs. The fans might be rotating slower to market it more quit.

As i do not have one i cant say for sure why the performance decreased but from my understanding of heat conductivity the fact that not all heat pipes don't touch the card does not reduce its cooling performance.

>>55687309
Why doesn't the GPU have a lid??

>>55687213
Bought the 970 strix a few months back, now my card rattles loudly when at load. Looked online and apparently it could be a fan wire that's not secured properly, how the fuck is asus a 'premium' brand?

So /g/ gimme the scoop, come on and tell me who the best manufacturer is this time for the x80 series. I gotta know what's hot.

>>55691878
>then

Let it go and go to bed. Didn't even seem to register with you that there's no comparison here with the 1080 cooler. They're completely different.

>>55687213
A year ago microcenter had a bunch of those asus direct cu ii cards real cheap open box. Grabbed one and ran hot but ended up bringing it back to upgrade my mobo/cpu instead. Glad I did, found this out after about the botched cooling, but I suppose if you slap an aftermarket cooler on it it doesn't even matter. But fuck ASUS for this shit

>>55691909
Heatspreaders are a relatively new trend in desktop computing, I believe introduced during the P4 era. CPUs and GPUs prior to then were all direct die contact. So it's still more common to see GPUs without heatspreaders.

And contrary to its name, it doesn't actually help with heat dissipation if your heatsink is any good. It's far more for physical protection of the CPU/GPU.

Many of the more hardcore enthusiasts actually de-lid their CPUs and mount their cooling setup directly with pretty good results. Some REALLY crazy enthusiasts even do direct-die water cooling.

Obviously it only matters if you're a really extreme overclocker for the sake of overclocking.

>>55691963
The 1080 cooler looks like they didnt weld the heatpipes together on some pictures. They might have lowered the amount of copper (total and alloy), worse fans, bad alu alloy and so on.

>>55692104
Which is irrelevant to the comparison I was talking about...

>>55692125
Then we can drop the fact that all pipes have to touch the die ?

>>55692197
Why should I? All you've said is "believe me"

>>55692209
Where ?

>>55692250
I'm summarizing.

>>55692261
Ok, from your point of view. Why is it bad for thermal performance when not all heatpipes touch the die ?
>>55692250

>>55692292
Same reason why a poor paste application results in lower temps.

>>55692329
Sure nigger.

>>55692044

I saw some comparison between direct-die cooling and with a heat spreader. It's only something like 2C cooler, you'd get better results by keeping the lid on and learning how to Liquid Nitro OC.

Anyone know where the fuck you can even buy a 1080 right now?

>>55689359
They are literally designed to do that.

>>55692811
>It's only something like 2C cooler
Nice meme. Delidding can lower temperatures by 20 degrees or now.

https://forums.overclockers.co.uk/showthread.php?t=18605867

File: 13823410_10153705603668244_1880488093_n.jpg (68KB, 540x960px) Image search: [Google]

68KB, 540x960px

they can barely get it right with only 3 heatpipes. asus 770 gtx cu II

>>55687213
>>55687315
>>55687426
So, these heatsinks are painted to help them retain heat, right?

>"ASUS's thermal solution uses five heatpipes; four of these directly touch the GPU's surface. You can also see a gray thermal pad that cools the voltage regulation circuitry. This is the exact same heatsink as on the ASUS GTX 1080 STRIX Gaming."

>Now we know why AIB 480 were delayed, we're getting coolers meant for nvidia cards!

WHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA

>>55694404
Neat bait. Here is a preliminary (You).

>>55694433
The other AIB partners haven't been known to do this like Asus does. That being said Sapphire is the only company for AMD pretty much.

>buying anything from Asus

They've been shit for a long time and you know it

Summer /g/ vs thermodynamics

File: laptop-reliability-1109.jpg (76KB, 600x350px) Image search: [Google]

76KB, 600x350px

>>55694541
0/10. Their motherboards and laptops have top tier reliability and had been for more than a decade.

>>55687309
Jayztwocents did a video on this a while back

It does not matter at all

>>55694617
>Jayztwocents did a video
>It does not matter at all

None of his vids matter

File: 1391142421364.png (129KB, 366x335px) Image search: [Google]

129KB, 366x335px

>>55693597
Never fucking mind he's using Liquid Metal, which has been know to drop temps by double-digits without delidding

so the thing i don't get right, a heatpipe is filled with a liquid that boils, rises and condenses again, but all these gfx cards their heatpipes boil at the heastinks, not the gpu? the lowest point is where the most heat is extracted, and all these heatsink designs have the lowest points somewhere far away from the actual core

>>55695782
oh i forgot, to check this, someone should do a comparison in temps between a regular case and a reverse atx case

>>55695782
It's just a slab of metal, Anon.

>>55687465
Wrong, it's the same design they use for everything so they can make it cheaper since they don't have to make an entire new assembly line for every single GPU.
It's actually cheaper like this.

So:
1. It doesn't affect cooling
2. It is cheaper

And you complain about it
If you look closer you'll see it's a heatpipe block rather than just heatpipes.

>>55694815
he delidded, replaced the TIM with MX-4 and pasted the IHS with liquid metal. You don't use liquid metal on bare die; you need abrasives to make contact with the surface and also the gallium metal is conductive.

>>55687257
Even if it seems like a redundancy the extra heatpipes will still conduct heat and cool it better than if there was only air in that space.

>>55695891
the pipes itself are filled with liquid, or else they aren't actually heatpipes. someone should crack one open to check

>>55694583
Got my ROG 'gaming laptop' for 5 years and it still runs fine. Basically has been on for 90% of that time, with me never even having cleaned it once.

Reminder that OP is a glorious faggot that can't into thermodynamics.

>>55698139
this

>>55687257
>Do you not see that only 3 out of 5 heatpipes would be in contact with the die?

Do you not see that the GPU die is cooled by the heatsink to the right, and the PWM chips and RAM chips are cooled by the heatsink to the left?

Where is the issue?