Sandia Develops Amazingly Efficient CPU Cooler

[jprahman]

People seem to miss the point that this cooler isn't for gaming PC and GPUs, but rather space constrained servers.

 

[ta152h]

The Pentium 4 is about to make a comeback. Between this and tri-phase transistors, they'd run at unheard of clock speeds.

And have roughly the performance of a Core 2 at 2 GHz .

 

[madooo12]

looks like a blower to me, I was thinking about using a blower for CPU cooling (something that looks like the intel stock cooler heatsink but rotates), too bad I didn't patent the idea, oh well...

 

[IndignantSkeptic]

I wonder if they can put this technology in smartphones and tablets because then those devices would be able to generate more heat which means their processors can run faster. Correct?

 

[JOSHSKORN]

Can't wait to see how this thing deals with overclocking! Oh, and me Tarzan want now!

 

[Guest]

I once got a gpu iwth a (plastic) fan simmilar to that, everything else was metal, and the first thing I thought was: why not a metal cooler(with tec pad), it would cool it down more wouldn't it?
So if you put those 2 on let's say a gtx 690 it could be clocked way higher without changing your cooler or switching to liquid cooling (some people still don't like it sadly enough).

 

[robot_army]

Would a quarter of the performance be 0.8 degrees Celsius per watt?
As heat transfer is per watt, ie if you have 10watt of heat energy to dissipate and 0.2 degrees Celsius per watt the temperature would rise by 2 degrees Celsius.

so 0.05 would give you 0.5 degrees Celsius far better

0.8 would give you 8 degrees Celsius far worse.

 

[freggo]

You know of course what the consequence will be... CPUs/GPUs that get hotter and use more juice; now that we have a way to cool 'em.

 

[vittau]

Now mix it with , and it'll be a whole new level for quiet cooling.

 

[vittau]

[citation][nom]vittau[/nom]Now mix it with , and it'll be a whole new level for quiet cooling.[/citation]
Dafuq happened to the link I posted, here it is: Noctua's active noise-canceling technology: http://www.youtube.com/watch?v=-atLQYr14zI

 

[behelit123]

Increasing surface area with a spiral design makes sense... But it seems like those gains are overstated. I've never used a Dynatron G950, but that thing is a huge 120mm fan heat sink. The Sandia looks like, as someone said, a hockey puck. Unless they have figured out how to bend the laws of thermaldynamics, it seems impossible that it is a "30x improvement in heat transfer". Honestly, that thing looks like it wouldn't beat a stock Intel heat sink.

 

[bribe]

Get this thing on Kickstarter so we can have it sooner

 

[Guest]

This thing is nearly two years old. This time last year, they were selling the commercial development rights. Surely a journalist would have enquired as to what the development status was? When will we see these things on the market?

 

[hetneo]

[citation][nom]bennaye[/nom]but how does the commercial dynatron cooler compare to today's most popular coolers such as the hyper evo and noctua lineup in general?[/citation]
You could google that.

http://www.xbitlabs.com/articles/coolers/display/dynatron-evo11-g950_7.html#sect1

Doesn't fare good against Hype 212 Evo 11 degrees difference. Looks like it performs similar to Intel's stock coolers.

 

[CrArC]

Would be nice combined with variable-geometry fan technology like those used in Apple's latest Macbooks. Very cool and quiet, I'd imagine.

 

[alyoshka]

This is actually petty old and this is the third time they've come up with this.

 

[behelit123]

The more I think about it, the more I realize this product is a fraud. I mean really? A heavy metal fan, with a fairly large air gap, blowing down on solid copper surface, no fins, is more efficient then a modern heat pipe based heat sink? Ridiculous. If the bottom piece works like a giant heat pipe, then where does the temperature differential come from? And that makes PERFECT SENSE putting the fan motor immediately above the hottest part of the surface, creating a dead zone. Seriously, who ever designed this needs to be fired. I can guarantee it won't work as predicted.

 

[puddleglum]

[citation][nom]alyoshka[/nom]This is actually petty old and this is the third time they've come up with this.[/citation]
Yep, and they still haven't shown us how they mount the thing. Comon guys, nice art work but show us the mounting brackets.

 

[msgun98]

[citation][nom]jamessneed[/nom]Instead of spinning the heat sink why don't they wrap a plastic shroud around it to force air through it properly. Something similar to stock GPU coolers so that it forces the air to circle around the cooler.[/citation]
That will end up much like a convection oven. You don't want the heated air stewing in a confined space, you want it the hell out of there.

 

[frank_drebin]

Any real-world benchmarks? for i7, i5? what about noise?

 

[pacioli]

[citation][nom]southernshark[/nom]Well I don't know if it will be revolutionary, but it will certainly be evolutionary. Hopefully this will assist in the needed "downsizing" of PCs.[/citation]
It is revolutionary... it spins!

 

[JamesSneed]

That will end up much like a convection oven. You don't want the heated air stewing in a confined space, you want it the hell out of there.

Obviously. Was thinking something like below, having the turbine type of fan sitting above the heat sink so the air would be pulled from the outside of the fins and exit via the center(The air flow would be reversed from a GPU well adn the turbine would sit right over the heatsink). Essentially pulling the air through the fin structure via the center. May work for mass production since its cheaper spinning mechanism.

http://www.tomshardware.com/reviews/radeon-hd-7950-review-benchmark,3207-2.html

 

[superp]

It is an amazing product, but I'd only buy it if the price makes sense.

 

[ramicio]

It's very loud. They need to quiet down the motor. I also don't buy the claims of stay dust free. Their logic is that it's because it moves. Well, normal fans' blades get dust stuck on them...

 

[behelit123]

Obviously. Was thinking something like below, having the turbine type of fan sitting above the heat sink so the air would be pulled from the outside of the fins and exit via the center(The air flow would be reversed from a GPU well adn the turbine would sit right over the heatsink). Essentially pulling the air through the fin structure via the center. May work for mass production since its cheaper spinning mechanism.

http://www.tomshardware.com/reviews/radeon-hd-7950-review-benchmark,3207-2.html

I don't see how this can work at all, actually. When looking at this, you have to ask yourself.... Where is the actual heat transfer taking place?

1. The thermal load interface. It's a copper plate, nothing new here.
2. They have basically taken a heatpipe (something that works best with a large temperature differential), made it really really short, and thicker then a Coke can. This CANNOT be as efficient as a regular heatpipe, because the heat has no where to go. It's literally just sitting there, right on top of the CPU.
3. A heavy, metal fan blows down on a smooth, copper surface. No fins, no contact area... Just this magic little fan pushing air down onto a smooth surface. Air. Think about that for a minute. Then think about the reason why they put fins on heatpipes and traditional heatsinks in the first place.

The laws of physics haven't changed, you still need surface area to radiate heat. Copper still radiates heat at the same rate it did before. I mean, this is pretty basic stuff here.