News Forget Water Cooling, Check Out This $7,300 Passive CPU Cooler

[Admin]

A seller on Etsy is attempting to commercialize a custom passive CPU cooler that spans from $1,800 to $7,300.

Forget Water Cooling, Check Out This $7,300 Passive CPU Cooler : Read more

 

[COLGeek]

Benchmarks, or it isn't real. Seems more like a piece of modern art than a usable component.

 

[ex_bubblehead]

I'm in full agreement with the Colonel. The images accompanying the story are obviously raytraces from Blender or such. And, with only one point of contact between each ball there's no way there will be much in the way of heat transfer. I call scam without even needing any numbers.

 

[hotaru.hino]

Did they individually hand solder all of those balls together or do they just stick there with some kind of magic?

 

[Friesiansam]

$7300... Lunatics with considerably more money than sense, form an orderly queue.

 

[Integr8d]

Benchmarks, or it isn't real. Seems more like a piece of modern art than a usable component.

Don't think about it. JUST BUY IT!

 

[COLGeek]

Don't think about it. JUST BUY IT!

You go first...

 

[tennis2]

Clearly fake, and yet..... gets a Tom's article.....

 

[Chung Leong]

In a pebble-bed reactor, convection flow is enough to keep it from melting down. There could be something to the design.

 

[ben_xman]

With every sphere in those stacks the transfer efficiency is horribly decreased. There is next to no surface area touching between each ball meaning after only a few cm, or heck, right there at the point that grouping meets the base barely any heat is being transferred.

Why offer a steel version? Are they unaware of how inefficient steel is at conducting heat?

Why 3D print the bases instead of just molding or milling it? Looks like the design would be far easier to mold or mill from stock than to spend the many more hours and expensive resources like filament and electricity 3D printing it. That's not going to add any sort of performance or structural benefit.

It's just a flashy design trying to be sold by looks (and wow, it costs so much upfront, so very Apple of us) without any real engineering thought put into it. I'd be surprised if it works any better than older intel stock coolers with the fan removed. Much less anyone actually getting one with how prototype it looks right now.

 

[USAFRet]

A couple hours in Blender/Rhino, and a day to print...I could make make something that looks exactly like that.

Show us one actually installed.

 

[spongiemaster]

A seller on Etsy is attempting to commercialize a custom passive CPU cooler that spans from $1,800 to $7,300.

Forget Water Cooling, Check Out This $7,300 Passive CPU Cooler : Read more

The i3/i5 version is a bargain at $305. Not clear how this one mounts to the socket either. Maybe super glue.

https://www.etsy.com/listing/827688854/passive-heatsink-for-i3-i5-cpu-desktop?ref=shop_home_active_2

 

[R_1]

A couple hours in Blender/Rhino, and a day to print...I could make make something that looks exactly like that.

Show us one actually installed.

thinking too hard, the spheres can be shaped into the top of a plastic bottle, looks like they were shaped in the top pf a plastic bottle.
spheres, glue, bottle and the tops are made,

 

[USAFRet]

Why 3D print the bases instead of just molding or milling it?

Because "3D printing" is cool and edgy.

"3D printed heat transfer base, heat exchanger based on pebblebed reactor design. Carbon fiber infused individual spheres specifically sized as to location and contact with all its neighbors, resulting in optimal thermal transfer."

A little more massaging of the text, and anything sounds cool enough.

 

[escksu]

If its a real product, there is only 2 materials that could work and so costly. Graphene and pyrolytic graphite. I am guess those grey balls are graphene - copper foam balls or pyrolytic graphite balls.

But one downside. Regardless of superior thermal conductivity, you still need adequate air flow, else it wont work

 

[TheHughMan]

Only while supplies last.

 

[COLGeek]

Only while supplies last.

But for a limited time, order one and get a second one free. Only pay a separate processing fee.

 

[cryoburner]

More importantly, it's unknown how the CPU cooler is fastened to the CPU socket on the motherboard.

Zip Ties.

This reminds me of that "Sketchy Heatsinks" series on Linus Tech Tips, where they attempt to make their own CPU cooler using various methods. Though this one may look better, assuming the renders are reasonably accurate and they don't just send you a melted blob of aluminum.

View: https://www.youtube.com/watch?v=TZHtlY6aDMw

View: https://www.youtube.com/watch?v=vL72lQ6dRZU

View: https://www.youtube.com/watch?v=3B1kjxO0sxM

 

[derekullo]

Top 10 Thermally Conductive Materials
Diamond
Silver
Copper
Gold
Aluminum Nitride
Silicon Carbide
Aluminum
Tungsten
Graphite
Zinc

https://thermtest.com/thermal-resources/top-10-thermally-conductive-materials

Obviously the best materials for a $7000 fan-less supercomputer heat-sink would be diamond balls suspended in a silver framework mounted to a TR4 socket.

Silver melts at 961°C and diamond melts at 4027°C in the absence of oxygen so this would technically be feasible.

Sadly neither diamond nor silver is organic, with silver even being antimicrobial.

 

[GenericUser]

I've never used etsy, so I'm just going to guess they must have absolutely awful return/refund policies for someone to want to risk selling frivolous garbage.

 

[digitalgriffin]

In a pebble-bed reactor, convection flow is enough to keep it from melting down. There could be something to the design.

Pebble beds only work where there is some medium to transfer heat. Air will not flow through this so there is no transfer medium

 

[Avro Arrow]

"There doesn't seem to be any heat pipes, though, which raises the question of how it transfers the heat to the balls. And we can't quite figure out what the term "organic" in the title means. "


On a more serious note, my guess is that the "organic" refers to the balls themselves which appear to be made of carbon with their non-reflective surfaces.

 

[rubix_1011]

If anyone can convince them to send me one to test, we can feature it on Tom's Hardware and see how it really performs.

 

[HotRod5353]

Only 7000 dollars? Can't someone come up with another cooler... not as good or as easy as water cooling , and sell it for 7 million dollars.

 

[Olle P]

In a pebble-bed reactor, convection flow is enough to keep it from melting down. There could be something to the design.

That's in line with my thoughts also, but with some caveats.

.., you still need adequate air flow, else it wont work.

That's one of the caveats. Probably won't be much natural convection through the pellet stacks.

Pebble beds only work where there is some medium to transfer heat. Air will not flow through this so there is no transfer medium.

It's possible to force air through. Pebbles are often used in filters for compressed air, so that shouldn't be a problem, but then if won't be fanless...

As for mounting: Screws with their heads on the other side of the motherboard holding the base down? We don't get to see the size of the base.

I can see this product as "possible", but with big questions regarding its efficiency, both in absolute terms and relative to price and weight.