Full Immersion Cooling

BrainEater

[H]ard|Gawd
Joined
Jul 21, 2004
Messages
1,087
Yep , seen it.

It's very nice work.
-----------
I'm sure taking the PR with a grain of salt.

100,000 pounds and 3 years for R&D ?

Epic Fail....I've published enough information on this forum alone to provide enough information for anyone to build a submersion rig....

:rolleyes:
 

peekama

Weaksauce
Joined
Jul 30, 2008
Messages
105
Thats actually euros. It comes out to about $145,307. And yea, the article makes everything sound a bit fancier than it is.
 
D

Deleted member 89137

Guest
What kind of liquid is that they're using? I can't understand the guys accent when he said it costs $100 / liter,,,
 

BrainEater

[H]ard|Gawd
Joined
Jul 21, 2004
Messages
1,087
heh , it's only 100$ a litre when you buy it used like they did.

You can get a deal on new FC770 for $2450 US for 3 gallons.

:D
 

BrainEater

[H]ard|Gawd
Joined
Jul 21, 2004
Messages
1,087
/me slaps persisting1.

As one of the two Creators of actual non-fishtank submersion cooling systems , That's not fair.

That's a fantastic , custom , submersion system .....It's an amazing piece of fabrication , even despite my laughing at the PR.

:D
 

hazexban

2[H]4U
Joined
Oct 20, 2004
Messages
2,525
Noob question, but I see the CPU still has the heat sink on it.

What would happen if they took it off and let the liquid be what is in direct contact with CPU?
 

Drakan290

2[H]4U
Joined
Sep 11, 2008
Messages
3,087
Noob question, but I see the CPU still has the heat sink on it.

What would happen if they took it off and let the liquid be what is in direct contact with CPU?

Good question.. I'd like to know the same.
I'd say that the liquid doesn't travel quick enough, so it'd overheat, but i'm probably wrong.
 

azuros

Gawd
Joined
Jan 1, 2008
Messages
785
Good question.. I'd like to know the same.
I'd say that the liquid doesn't travel quick enough, so it'd overheat, but i'm probably wrong.

That is exactly what would happen, the only way i would see it as a viable option would be to use iso with dry ice cubes but that would be forcing a constant way below ambient temp. at the very least attaching a heat sink to it and some type of flow mechanism would work with it.
 

hazexban

2[H]4U
Joined
Oct 20, 2004
Messages
2,525
Good question.. I'd like to know the same.
I'd say that the liquid doesn't travel quick enough, so it'd overheat, but i'm probably wrong.

My theory though is the speed of travel for the liquid doesn't matter, it's the ability of the liquid to conduct the heat. The HS will still heat up and pass the heat on to the liquid but there is a conductive loss from the CPU through the HS to the liquid.

I haven't taken a physics class since college but I think this would work better without the HS.

Anyone able to give scientific back up to why the HS being on works better in the liquid?
 

BrainEater

[H]ard|Gawd
Joined
Jul 21, 2004
Messages
1,087
That is exactly what would happen, the only way i would see it as a viable option would be to use iso with dry ice cubes......


:D :D

-----------

Sure hazexban.

I can tell you exactly what happens.

You get localized boiling because of the high watts/cm2 , which disrupts the heat conduction to the liquid , thus causing problems.The heatsink reduces the thermal density (increased surface area) , allowing a greater conduction to the liquid.

Thermal transfer operates the same way regardless of whether it's in gas or liquid.Remember , gas=really thin liquid.

:D
 

hazexban

2[H]4U
Joined
Oct 20, 2004
Messages
2,525
:D :D

-----------

Sure hazexban.

I can tell you exactly what happens.

You get localized boiling because of the high watts/cm2 , which disrupts the heat conduction to the liquid , thus causing problems.The heatsink reduces the thermal density (increased surface area) , allowing a greater conduction to the liquid.

Thermal transfer operates the same way regardless of whether it's in gas or liquid.Remember , gas=really thin liquid.

:D

thank you sir! good luck with your awesome project!
 

Cerulean

[H]F Junkie
Joined
Jul 27, 2006
Messages
9,477
Old news (to me anyway) -- appeared on Slashdot a few months ago.

But yeah -- that stuff is very interesting! :D I would like to play around with one and closely examine it, as most people here would.

Very unique!

I'd say that the liquid doesn't travel quick enough, so it'd overheat, but i'm probably wrong.
It depends on how well/quickly the heat from the die core can transfer to liquid, versus from the die (through thermal paste; composition of thermal paste matters) to a solid (metal material matters too).
 

sl1982

Limp Gawd
Joined
Jul 24, 2008
Messages
365
Haha i just looked at the wiki on flourinert. You could have some lab rats swimming around in the stuff. I guess they would die when they get hungry though. Not to mention feces... ok bad idea, nevermind.
 

ADDICT76

Limp Gawd
Joined
Jul 1, 2008
Messages
169
The screen savers did this yearssssssssss ago and ran theirs with no CPU cooler. It boiled around the processor but never overheated. Though the stuff evaporated at the rate of like a gallon a day!
 
Top