News DIYer Passive Cools Core i9 with 8-Pound Copper Block

[Sniper29a]

Drill holes, or mill slots to increase surface area...increases cooling effect, thermal bridge efficiency..it may work...so much weight per cm2 of MB must be brutal.

Servers, workstations use passive coolers for years ;-) they just increase surface area by thin fins

 

[bit_user]

Servers, workstations use passive coolers for years ;-) they just increase surface area by thin fins

It's slightly misleading to call them passive, since server heatsinks depend on high airflow from a bank of fans that feed directly into them - they just don't happen to be directly attached. The servers I've seen even have airflow guides to help force the air from those fans through the heatsinks.

 

[Sniper29a]

It's slightly misleading to call them passive, since server heatsinks depend on high airflow from a bank of fans that feed directly into them - they just don't happen to be directly attached. The servers I've seen even have airflow guides to help force the air from those fans through the heatsinks.

Because passive cooler doesn't work itself. At certain point, heat doesn't spread further.

If you have air around the passive block, it remains hot. It won't spread in whole room ;-)

You have to always force air; otherwise efficiency would be crap.

I used to have large Alu plate with 5x GPUs on it...plate was warm about 5-10cm from heat source...that's it. It won't spread to the whole plate ;-) I am guessing it something like resistance or so...no idea...I just did CAD simulation.

It is simple nature of seeking opposites...in this case warm-cold

You see it from this example...CPU idles...no issues...capacity of near copper material sinks heat at same rate...you add 100W...you are done. Whole block won't suddenly get hot, sinking all heat from CPU ;-)

Efficiency is linear with temperature delta/difference. It may work if ambient temp is 0°C or below 0.

 

[InvalidError]

Because passive cooler doesn't work itself. At certain point, heat doesn't spread further.

It can when everything is setup to facilitate natural convection as much as possible and you don't mind the huge heatsinks required to get there on larger loads. The good old CM212+ can passively keep up with about 60W of load when mounted fan-side parallel to the table in an open-air bench if you don't mind the CPU being around 80C.

 

[Sniper29a]

It can when everything is setup to facilitate natural convection as much as possible and you don't mind the huge heatsinks required to get there on larger loads. The good old CM212+ can passively keep up with about 60W of load when mounted fan-side parallel to the table in an open-air bench if you don't mind the CPU being around 80C.

yes, it can work if you have large surface...such as fins commonly found on radiators, passive heatsinks.

A solid block of metal won't be so efficient. You can put 10kg of solid copper on CPU...it will have always ~80°C because it won't be so efficient on the other end of solid material.

For example, 4kg as we see here has 80°C. 10kg may have 75°C...then it rips of socket from MB ;-)

It is simple physics, thermodynamics or whatever it is called.

I have Corsair H150i...it can silently sink 120W @ <60°C...AV512 kills it, though. Imagine how large passive cooler would you need for Ryzen 5950X @ 4.3GHz ~110W

This example has similar power output, and cannot keep up

 

[InvalidError]

Imagine how large passive cooler would you need for Ryzen 5950X @ 4.3GHz ~110W

Not really that large: 110W is only 2x 60W which my 212+ can handle passively, so you'd only need something about twice as big. The real challenge is moving heat away from the core and distributing it evenly throughout the entire fin stack fast enough as the speed at which heat can be moved around is the second biggest limiting factor of conventional HSFs after form factor constraints.