What are closed loop water coolers for if they perform worse than air coolers but cost more?
>>51724504
So people can act like they have a custom water cooling setup without actually making a full loop.
It's about keeping temperature fluctuations to a minimum amplitude. You'll notice that idle and load temperatures are at a smaller delta than air. Pure air setups can be a lot louder at times than closed loop.
>>51724665
>Pure air setups can be a lot louder at times than closed loop
And CLCs can be a lot louder than any decent air cooler 100% of the time. There is no CLC in existence that can offer the noise levels of a huge dual tower air cooler with decent fans. Pump noise is unavoidable.
>>51724504
d15 looks ugly as fuck
>>51724705
All CLCs look ugly as fuck.
>>51724504
So people don't warp their motherboards and cpus
>>51724752
>things that never happened because of weight
>>51724771
Intel, please leave.
>>51724800
I said weight. Intels shit happens because the whole assembly is too tight
>>51724665
>Pure air setups are loud
My D15 begs to differ.
>>51724705
If you don't have a windowed case this a literal non-issue.
>>51724504
For higher end chips. No air cooler can handle an overclocked x99 chip.
>>51725015
Air coolers can absolutely handle it.
I have a 4790k at 4.7ghz with a D15. Idle is 26c, when gaming it's very rarely above 45c. Even when I have my cpu held at 100% use accross all core and hyperthreading during photo rendering it tops out at about 60c.
>>51725076
Your mainstream shit isn't x99 faggot
>>51725219
The x99 chips aren't that much more extreme or hot running, and if you think they are you need to do some research.
>>51725219
Are you fucking high? Plenty of people are running air cooling on high wattage chips, X99 isn't some mystical fucking volcano chip. An example is the FX lineup, they use as much juice and in some instances a LOT more than X99 chips, yet they are very docile with decent air cooling. The performance of the FX line might be lacking compared to X99, but they are basically the same thermally.
>because it makes it easier to get to components and looks neater. Also quieter for the same amount of cooling.
To make money off of gamebros.
>>51725451
This desu senpai
Nothing is blocked with a water cooler.
>>51725574
>I want to use gaming memory with 5 inch heat spreaders
>>51725906
I like to check things that go wrong easily.
Same reason I don't super zip tie every cable in my life.
Ease>everything else
They're baller
An AIO will always look cooler than a big heatsink
>>51725351
lol
>>51725382
nope
>>51724504
rice
autism
>>51724705
like this guy
>>51725936
>ease
There are more parts that can break on an AiO tho. Pump, built in res, tubing comes off and leaks all over the place. Unlikely, yes, but it can happen.
A heat sink is literally just a chunk of metal with no moving parts other than the fans. If this is not the easiest solution, I don't know what is
>>51725015
4.3GHz core / 4.3GHz uncore / 1.2Vcore on a 5960X using a Noctua NH-D15.
Temps never go above 75c and fans never go above 1600rpm.
>>51724665
>d14 at idle
fucking nothing
>d14 under full load
fucking nothing
maybe next time don't use some $20 air cooler or some POS stock cooler
https://www.youtube.com/watch?v=JT_u_eLVoIc
>YFW WATERCOOLING KEKS
AYY
money.
I can't believe >>51725494 is the only post that understands this. The rest of you can't be that stupid.
>>51724504
Because the CLC's you have in your graph are Corsair garbage. There are good CLCs from Swifttech and EKWB that outperform Air easily.
Actually I'd like someone to tell me how the EK Predator is any worse than a custom loop. It's easily expandable.
>>51724504
rice, duh
>>51726294
I would hardly call the EK predator closed loop.
>>51725451
It definitely isn't quieter.
>>51726397
Well it's an AIO in the same category then.
>>51726294
Provided some evidence..
Also, those swiftech and ekwb kits aren't exactly CLC. They are literally assembled custom loops.
>>51726075
run prime95 and post results.
>>51726294
>I'd like someone to tell me how the EK Predator is any worse than a custom loop
it's got a shitty block with a shitty ddc
you can literally build a real loop for less.
So far in all of these CLC threads I've seen the only valid point has been that it looks "baller", which is highly subjective.
This is, of course, ignoring the fact that a large portion of /g/ prefers their computers to look like a featureless black monolith with no way to see the innards, making a CLC something you'll only ever show off in guts threads.
>>51725076
You idle using an air cooler at 26C?
doubt.jpg
>>51726931
You never know, he could live in an igloo.
>>51726481
It is as long as you don't get Corsair garbage with their garbage 2700rpm fans. Problem is people think Corsair is the only one that makes AIOs and to a lesser extent CM. Corsair garbage IS ricer shit that is worse than air but all other AIOs get put in the same bin of shit and it isn't fair.
>>51726527
>They are literally assembled custom loops.
There's literally no difference in design between Corsair's AIOs and a H220 except you can expand a H220 since they use fittings that you can reseal.
See pic, also this is with Swifttech fans, stick some GTs or EK Vardars then you got something even better.
>>51727107
Also another thing to note is the D15 is using 140mm fans, if we compare it to a H240x then the gap would be even wider.
>>51727107
>but all other AIOs get put in the same bin of shit and it isn't fair.
>After posting a chart showing every other CLC to be very loud and performing unconvincingly
You're not making a very good point here mate.
>>51727263
They're complete no names though. Be honest, 90% of all the builds with a AIO in it is a Corsair and the last bit might be a CM or NZXT, you hardly ever see anything else. Swifttech and EKWB are known brands for water cooling.
>>51724504
The point of a water cooler is providing a decently large surface area for heat transfering(the radiator) availble for mounting somewhere else other than right on top of the CPU and then using a fluid to transfer the heat load from the CPU to the radiator using a pump.
This is great if you don't have huge amounts of space imediately on top of the CPU, but have it somewhere else availble in your case.
Over short distances, using a solid material would be better for conducting heat, so in that case an air cooler would provide for faster transfer and, given a nice air flow, better performance.
However, like >>51724665 said, because water has a higher specific heat than copper or any other metal, you also keep temperature fluctuations to a minimum. That way the water cooler setup gets hotter slower than the air cooler.
All this said, closed loop water coolers are meant to facilitate and make having a water cooling setup for the reasons described above easier for the masses.
>>51727322
I'm sorry but you've just gone from
>Everyone who isn't Corsair are getting a bad name because of Corsair
To
>Everyone is shit except these two names
That's quite a far cry from your other point. Regardless, the use case for a CLC of any kind is pretty small unless you're aiming for a very high end x99 chip. If that's the case, why not just got for a custom loop which will perform considerably better again? If you're going to buy a CPU approaching a thousand dollars then you certainly should consider a custom loop.
i obviously don't go pulling my hsf off all the time, but over the years i've grown tired of fucking with xbox huge tower coolers with complicated mounting solutions. clcs these days are very easy to install and do their job well enough.
>>51727392
Maybe you should read, yes all other AIOs get put into the garbage bin including the ones that actually are garbage. I never said all other AIOs are top tier dumbo. People only know AIOs by the Corsair name is all I'm saying.
>the use case for a CLC of any kind is pretty small unless you're aiming for a very high end x99 chip.
You couldn't be more wrong. x99 doesn't have as much of a heat problem because Intel actually solders those chips. I own both a 4820k and a 4770k. 4820k doesn't go over 70 with a 4.5Ghz OC, the 4770k however has a better cooler and goes over 80 at full load with the same overclock. Both have similar voltages as well. It's very needed for Haswell pre devil's canyon and non-E Ivy Bridge.
>If that's the case, why not just got for a custom loop which will perform considerably better again?
Because that requires more effort and also you can get a H220X for $140 which is less than most loops. The H220 went on sale for $100 for a good while.
>>51727490
You said, and this is copied from your post
>all other AIOs get put in the same bin of shit and it isn't fair.
The bit about "all" is a pretty stark contrast to "everyone except these two brands are shit"
>Haswell needing more cooling than x99
But that's just wrong, go look at the wealth of benchmarks not only on the web as a whole but even just in this thread. Even the humble 212 evo has no problem keeping up with average CLCs several times its price when it's handling the lower TDP of an Haswell CPU. The only charts that the CLCs pull ahead in any way are the ones using a CPU with a TDP greater than 120 watts.
As for the thermal paste buggery having an input, it really doesn't. It's a simple equation, energy out has to be equal to energy in or else the temperature will rise until there's an equilibrium. That's why we look at the temperature delta and not just the temperature, because the chip used in the benchmark affects the highest temperatures. It's also why they use the exact same chip for all the tests, because even between two identical model numbers the difference is enormous.
>Price on the h220
Again, if you're buying a CPU approaching one thousand dollars, yes, that's $1,000, then you'll be more than likely to have the money to spend on a decent water cooling solution. As for convenience, I'll concede that point to a certain extent with the note that if you're finding water cooling difficult you'll probably be needing assistance for the entire rest of the build anyways.
>>51727733
I'm not going to repeat myself, I already told you that isn't what I meant by that. Also you're strawmanning since I never said anything but those brands, those are just two that I know that perform well. CM also has a cooler that's basically a h220.
>But that's just wrong, go look at the wealth of benchmarks not only on the web as a whole but even just in this thread.
See pic
>As for the thermal paste buggery having an input, it really doesn't.
It really fucking does see pic and also see the next pic I post
>Again, if you're buying a CPU approaching one thousand dollars,
You don't need to spend 1k for x99 and this is moot anyway since non-E is hotter.
>>51727916
>>51727733
Here's the difference in temps using better TIM on Skylake
Also here's an article on it for Devil's Canyon
http://www.kitguru.net/gaming/uncategorized/anton-shilov/intels-devils-canyon-chips-ngptim-is-still-not-efficient-research/
I have to go out now but I might respond in an hour or two.
>>51727916
>I'm not going to repeat myself
And yet you're going to make me repeat myself
>It really fucking does see pic and also see the next pic I post
Again, you look at the temperature delta, not the temperature. The thermal paste means that haswell runs inherently warmer, regardless of cooler used. The lower TDP, on the other hand, means that exactly which cooler used is not going to make as big a difference because there's less thermal energy to be moved. Go look at the graphs in this thread. See how there's a huge advantage to the CLCs compared to the lower air coolers on the high end CPUs? Now look at the ones with mainstream CPUs. See how even the 212 is keeping up? Thermal energy is the deciding factor here.
>You don't need to spend 1k for x99 and this is moot anyway since non-E is hotter.
You're right, I'm taking the extreme end of the scale. Regardless, none of the x99 processors are cheap and the added costs associated with the enthusiast chipset and so on and forth still make it far more reasonable to consider a custom loop.
>>51727961
The difference in TIM does not have any impact on the matter at hand, it STILL COMES DOWN TO TDP.
We had HUNDREDS of these threads when the issue was revealed. We had EVEN MORE when the high temps on AMDs 290X reference cards were revealed and people thought they were heating the room up way more because they were running hotter. The absolute temperature is irrelevant, it's the thermal energy that is important. How small do I need to make that teaspoon before it will sink in?
>>51726756
Already did when I built the machine. As I said, never goes above 75c.
>>51726931
My CPU idles at 15C and my GPUs in SLI idle at 20C.
>Sweden winrar
>>51724504
They were always for meant for silence. I'm not sure why people on /g/ think they're cooler than traditional heatsink / fan.
>>51724705
Fuck off it does.
Excuse the cables and tape, I'm waiting on a few things before I properly mod this case.
>>51728535
>They were always for meant for silence
You don't add more moving parts to achieve silence. If you want silence you either get a Noctua and put the fans as low as possible or you get one of those laughably large passive heatsinks.
>>51728597
I know appearances are subjective but there is nothing pretty about their stuff. The choice in colour is bland to the point of being offensive, the shapes are clunky but not enough to be exciting and the half arsed attempt at making them look decent just ends up killing any utilitarian appearance they could have achieved.
That being said they're amazing an anyone who discounts them because of their looks should be shot on the spot, repeatedly.
>>51728052
>>51728114
We're not arguing TDP or if the E-line produces more heat here dummy. Of course it does but because of the superior heat transfer you aren't limited by temps nearly as much as you are with the mainstream K chips.
> See how even the 212 is keeping up?
It's 90c in the OP... I don't even see it on the one I posted which is the only one with a E chip. You're going to have to point out the differences to me. Anyway this case might be true but it still misses the entire point.
>The difference in TIM does not have any impact on the matter at hand
And here is where the whole issue is, TDP isn't the only factor here bud. A cooler that isn't transferring heat as efficiently is going to need to work harder in order to cool. It's that simple. How small do I need to make this teaspoon?
>>51729494
>A cooler that isn't transferring heat as
But the cooler is, the IHS isn't. Temperature is irrelevant, thermal energy is important.
>It's 90c in the OP... I don't even see it on the one I posted which is the only one with a E chip
Way to ignore the one thing I've been singing repeatedly. Once again now, sing the chorus with me;
Absolute temperature is what now?
>>51731302
>But the cooler is, the IHS isn't.
Semantics, the result is the same. The cooler needs to work harder.
Sing all you want, you droning on about TDP doesn't change anything I said. You can't even point out the baseless claims you made.
>>51725015
>No air cooler can handle an overclocked x99 chip.
Huh? C/W is C/W dude. Linear fucking relationship.
>>51727107
>no CM nepton 280L
every fucking time
>>51731549
>The cooler needs to work harder.
No, it doesn't. It's dissipating LESS heat. What hit of "temperature is irrelevant, thermal energy is what matters" is difficult. Less thermal energy, less work for the heatsink regardless of the absolute temperature. That is why you see a closer grouping on the mainstream, lower TDP products than you do on the enthusiast grade power guzzling monsters.
Like I've said earlier, we had this conversation when the 290x came out. Temperatures have almost nothing to say on the matter, you're Tring to get rid of thermal energy here.
>>51732127
We've already gone over this and all you've done is repeat yourself instead of actually making a rebuttal. Yes, we've already gone over that it has less heat output you bonehead. What part of loss of heat transfer don't YOU understand? It doesn't matter if it has less to work with if it's not working efficiently in the first place.
Temperature is not irrelevant, these solderless chips hit TJ max far easier than any soldered E processor. I don't know why you can't get this when I even posted goddamn proof of it but that still isn't enough for your dense head. Keep droning on. I'm going to bed.
>>51732127
Oh and you haven't even pointed out this grouping either, the Hyper 212 is in only one of these graphs
>>51732240
The prof you've posted yourself contradicts you and shows my point. The grouping is there.
>Bonehead
Nice, can't make a point so you go to insults.
>Going to bed
Sleep well anon, hopefully you'll have a think on this after you wake up.
>>51732258
Hyper 212 was mentioned as an extreme edge case, as it's often trotted out as the epitome of price/performance and more importantly is absurdly cheap.
The point stands with all other air coolers, especially those more expensive.
>>51728597
fucking nasty shit, pls fuck off.
Rad fans act as exhausts as well
>muh colors
Non-friendly reminder that Noctua actually makes grey but also black fans.
>>51733214
>muh premium price and marketing
Non-friendly reminder that Noctua actually doesn't make the best fans and EK Vardars are superior.
>>51733988
> EK Vardars are superior.
I just need two 120mm fans that can dissipate a gpu sucking down 375w.
looks
>recently learned that Cooler Master is selling shittier versions of Hyper 212 EVO in Europe compared to the rest of the world
>fan maxes out at 1600rpm instead of 2000rpm
I feel raped financially, mentally and thermodynamically.
>>51733988
Noctua's fluid bearings are very reliable and last many years without getting noisier. EK uses ball bearings which often get noisy as they age, and the specified 50000 hour MTBF is not very impressive.
>>51734369
My GT's with ball bearings sound just as good as the day I bought them. Been over 5 years now. I even have two unused ones to compare.
Thing about ball bearings is you can easily relubricate them if issues arise. Noctua's fans require a seal so you're kinda fucked, some people have reported leaks. Only time will tell if Noctuas will last, their warranty is very good though. Don't think anyone has them beat there. I think only Noiseblocker matches it.
>>51726931
Hyper 212 evo, 3 minutes of idle, fans at 7v (400-600rpm)
Check those overnight low temps.
Akasa Venom Voodoo, how good are these?
>>51732910
>>51732910
What am I supposed to think about when you didn't even make a rebuttal here? Nor did you even point out something simple to have a basis for your claims. The one that you did point out fell on your face. What are the other air coolers? Point it out senpai. Looking back this really is funny how baseless all this is.
Again, you have these solderless mainstream K chips hitting TJ max with mild OCs even with high end AIOs. You run into a temp issue before you do with soldered E chips which can get 50% OCs before running into temp issues. But yes, the shitty TIM doesn't matter right? TDP is everything and the soldered E chip needs the better cooler regardless of facts.
