Cooling Wars: Episode IX - The Rise of Air v Water

[xtreemchaos]

i used to use Acrylic its fragile it cracks easy its not as good a PETG to bend but it do look better it has the look of glass. ill still use it if a customer asks but for my self no not anymore ive moved on .

 

[WhitetailAni]

The last bit I need for P4HT's water loop is 2 80mm radiators with 1/8" inner diameter pipe (It can be 1/4" but adapters aren't cheap).
Any thoughts on which radiators I should get? I'm not looking to go ultra cheapo mode but I don't want to spend $70+ per radiator. I'm using Arctic F8 80mm fans in a Lian Li PC-62 if you're curious.

 

[TheinsanegamerN]

I don't disagree with that, but reviews for cheap products are scarce. It's crazy how hard it is to find a review of say Pentium G6400 or Celeron G5900, compared to i9. It's also hard to find thermal tests done on cheap cards like 1650 from say Asus Phoenix. And yet pretty much every RTX 3070 and up are tested in depth. Same with motherboards and SSDs. If you want to build a cheap machine, then finding reviews for components is next to impossible. Even worse if you want to learn something about regional brands like Alpenfohn, SilentiumPC or Chieftec.

See, as someone who has used a recent pentium product in a build, I want to see someone watercool a pentium. I know from experience that the pentium will run without a heatsink. Not fan, heatsink, just bare IHS is sufficient in a case with no fans for some time before it overheats with heavy constant load. i want to see someone put a waterblock on it and keep it at room temp, for lulz.

 

[AusWolf]

See, as someone who has used a recent pentium product in a build, I want to see someone watercool a pentium. I know from experience that the pentium will run without a heatsink. Not fan, heatsink, just bare IHS is sufficient in a case with no fans for some time before it overheats with heavy constant load. i want to see someone put a waterblock on it and keep it at room temp, for lulz.

Though not a pentium, I put a 120 mm AIO on a 35 W T-series Sandy i3 once. It ran under 50 °C in every scenario... without a fan.

 

[Valantar]

Somebody by the name of, 'you,' said you ran your water cooler that hot. Anyway, you don't determine biofouling, Dow Chemical Company does. That is where the biocides are produced.

Finally a straight answer! So that's where you claim to get that from. Except that I said this:

I have a custom loop, and run my CPU around 80°C.

So, what does that say? That my CPU runs at around 80. Not the cooler. Not the water in the loop. The CPU. This was my entire point: it's clear that you flat-out misread what was said, and took that in a rather weird direction. All I was trying to say is that nobody is claiming their water loop runs at that temperature, nor is anyone claiming that this would be safe - or even possible at all. You can't run a water loop at 80°C liquid temperatures. Nobody has said they did so either. You misread what I said. Discussion done. Now can we please move on from this stupid side-track?

 

[mtcn77]

So, what does that say? That my CPU runs at around 80. Not the cooler. Not the water in the loop. The CPU. This was my entire point: it's clear that you flat-out misread what was said, and took that in a rather weird direction.

You seem to forget water is 10 times better at extracting heat than copper. If your cpu is 80°C, your copper cold plate could be 79°C and water could be 78.9°C while still at the same carrying capacity, that is the difference. Your water is not far behind.

 

[MxPhenom 216]

That's an example of what people are afraid of with liquid cooling.

Acrylic has other problems. I generally recommend soft tubing.

Like what? soft looks like shit.

 

[AusWolf]

You seem to forget water is 10 times better at extracting heat than copper. If your cpu is 80°C, your copper cold plate could be 79°C and water could be 78.9°C while still at the same carrying capacity, that is the difference. Your water is not far behind.

Watercooling works with much bigger temperature differences than that. Water that hot would never be able to keep your CPU anywhere near its operating temperature range, exactly because of the transfer among materials that you mentioned.

(Not that I understand what the point of the argument is, but never mind.)

 

[mtcn77]

Watercooling works with much bigger temperature differences than that. Water that hot would never be able to keep your CPU anywhere near its operating temperature range, exactly because of the transfer among materials that you mentioned.

(Not that I understand what the point of the argument is, but never mind.)

I don't think you really understand specific heat, but carry on.

 

[Valantar]

You seem to forget water is 10 times better at extracting heat than copper. If your cpu is 80°C, your copper cold plate could be 79°C and water could be 78.9°C while still at the same carrying capacity, that is the difference. Your water is not far behind.

Uh... No. That's not how thermal transfer works. You're assuming essentially perfect and immediate transfer. That is a semi-utopian best case scenario. A real water block is never that good. Also, I have temperature sensors in my loop, so I know how hot the fluid gets. Now, I don't have one directly after the CPU, and some heat does sink into the GPU when it's idle, but... My fluid temps peak in the low 40s when the entire system is running flat out. Yeah, no, the GPU is not lowering water temperatures by 40-50 degrees (especially seeing how the radiator only lowers temps by 2-5 degrees depending on pump speed). So no, at no point in my loop is the fluid anywhere near 80 degrees.

Edit: derp, 2-5, not 2-50.

 

[TheoneandonlyMrK]

/poll

Seconded.

 

[AusWolf]

I don't think you really understand specific heat, but carry on.

I never even mentioned anything remotely related to specific heat. I was talking about heat transfer among materials, but carry on.

 

[MxPhenom 216]

Uh... No. That's not how thermal transfer works. You're assuming essentially perfect and immediate transfer. That is a semi-utopian best case scenario. A real water block is never that good. Also, I have temperature sensors in my loop, so I know how hot the fluid gets. Now, I don't have one directly after the CPU, and some heat does sink into the GPU when it's idle, but... My fluid temps peak in the low 40s when the entire system is running flat out. Yeah, no, the GPU is not lowering water temperatures by 40-50 degrees (especially seeing how the radiator only lowers temps by 2-50 degrees depending on pump speed). So no, at no point in my loop is the fluid anywhere near 80 degrees.

mtcn is our resident HVAC expert...or so he thinks

Uh... No. That's not how thermal transfer works. You're assuming essentially perfect and immediate transfer. That is a semi-utopian best case scenario. A real water block is never that good. Also, I have temperature sensors in my loop, so I know how hot the fluid gets. Now, I don't have one directly after the CPU, and some heat does sink into the GPU when it's idle, but... My fluid temps peak in the low 40s when the entire system is running flat out. Yeah, no, the GPU is not lowering water temperatures by 40-50 degrees (especially seeing how the radiator only lowers temps by 2-50 degrees depending on pump speed). So no, at no point in my loop is the fluid anywhere near 80 degrees.

It doesn't matter exactly where you have the temp sensors for the fluid in the loop. After a couple minutes, it should reach equilibrium and water temps will be the same throughout. I have one sensor in mine as a plug for an unused port on my res. My water temps bounce around 38-41c depending on the game.

 

[mtcn77]

I never even mentioned anything remotely related to specific heat. I was talking about heat transfer among materials, but carry on.

You must be a different type of science enthusiast, but do carry on.

 

[ThrashZone]

You seem to forget water is 10 times better at extracting heat than copper. If your cpu is 80°C, your copper cold plate could be 79°C and water could be 78.9°C while still at the same carrying capacity, that is the difference. Your water is not far behind.

Hi,
Stagnant water/ fluid maybe but we all have pumps that circulate fluid so your thinking/ science does not apply here.

 

[AusWolf]

You must be a different type of science enthusiast, but do carry on.

What's your point? Let's not sink to the level of making things personal.

 

[mtcn77]

Hi,
Stagnant water/ fluid maybe but we all have pumps that circulate fluid so your thinking does not apply here.

Really, like equilibrium? That is what I said. Heat exchange is the same at the radiator, the ihs and the water.

What's your point? Let's not sink to the level of making things personal.

Heat transmittance function is not up to debate.

 

[ThrashZone]

Really, like equilibrium? That is what I said. Heat exchange is the same at the radiator, the ihs and the water.

Hi,
Yeah you worded your statement earlier a little freaky but i get what you're saying now my bad.

 

[The red spirit]

See, as someone who has used a recent pentium product in a build, I want to see someone watercool a pentium. I know from experience that the pentium will run without a heatsink. Not fan, heatsink, just bare IHS is sufficient in a case with no fans for some time before it overheats with heavy constant load. i want to see someone put a waterblock on it and keep it at room temp, for lulz.

Well, it might even run well under all loads with just stock heatsink without a fan. It would be nice to know that.

 

[mtcn77]

Hi,
Yeah you worded your statement earlier a little freaky but i get what you're saying now my bad.

I never take an apology for granted. Eventually, we shall all agree what is happening is in accordance with physical phenomena and taking the optimal heat transfer as the benchmark is a good starting point to which we can add side functions.
Enough trolling, good night gentlemen.

 

[lexluthermiester]

Like what?

Oh, like...

its fragile it cracks easy

...this for example.

Big problem when you want to keep a sealed fluid system actually sealed. But it also tends to react with the chemicals used to kill bio-growth, which causes discolouration.

soft looks like shit.

That is YOUR opinion. A lot of people would disagree and consider hard tubing a very real PITA, especially if you do frequent upgrades.

 

[Bones]

As promised earlier.

Still have some things to fix and refine but the basic concept of it's application works.

 

[Space Lynx]

As promised earlier.

Still have some things to fix and refine but the basic concept of it's application works.

this is going to be epic. honestly can't wait to see what the final product looks like. also this is kind of genius, I am surprised the big companies have not thought of this yet.

 

[Fangio1951]

As promised earlier.

Still have some things to fix and refine but the basic concept of it's application works.

hi m8,

Nice bit of DIY, but, where's the cooling component of the loop = no radiator/fans ??

 

[Space Lynx]

hi m8,

Nice bit of DIY, but, where's the cooling component of the loop = no radiator/fans ??

I am confused as well, I was thinking he was 3D printing then like shaping a connector/enclosure to support an AIO mounting device... but I can't tell from the video either what is going on really. Wither way, it is a great idea and neat concept.