>The fluid in this server is boiling at 56 degrees celsius, cooling it more efficiently than traditional H2O-cooling.
Explain.
Explain what, cuck-boy
hurt to read that
>>58363925
when liquid changes phase, it draws off a large amount of energy in the form of heat.
This liquid is nonconductive and changes phase at a point well below normal operating temperatures for computers, making it more ideal than water for cooling a computer. It might have less thermal capacitance but can move the heat energy around more efficiently.
>>58363967
Thanks, lad.
>>58363967
But the liquid itself need to be cooled. Is it a loop to a radiator we dont see?
>>58364569
>>58365178
this is why I come here. pretty neat.
>>58363925
>>The fluid in this server is boiling at 56 degrees celsius, cooling it more efficiently than traditional H2O-cooling.
boiling liquid and condensing vapour have the largest convection coefficient (followed by pure liquid). as the liquid boils, it rises and discharges heat to the surrounding air. as the heat is released, it condenses and goes back to the bottom. it's the same way heatpipes work (liquid evaporates and carries heat, then condenses again)
>>58363925
what fluid is that to not cause it to electrocute?
That's a pretty cool setup
>>58365252
It's neat for a 250+kW Bitcoin mining rack. Where you can ditch the heat spreaders and put asic boards very close to each other.
But for regular x86 servers it is not so great. At least for now.
>>58365334
3M Nover 7100 or similar. Expensive as fuck.
Sure is fucktard in here.
>>58365303
the bottleneck is still the condensation surface, i would not recommend unless you want to enjoy some funny fumes or a time bomb. The hardware would be thermally safe while there is still liquid tho.