Sandia-designed CPU cooler lacks fan, rotates heatsink instead

[DeltZ]

I'm a bit skeptic about this fan. As it appears not to contact directly to the CPU's IHS then than means we'll get a slim layer of air between them. Air is a VERY bad heat conductor so the hypothetical efficiency you gain with this fan's design is annulled by the heat conductivity.

Anyways, even if it works good, I doubt this could be cheaper and more effective than a Corsair H100 with Collaboratory Liquid Ultra TIM.

part of their basis theory is that air conducts a lot better when it's being sheared by the continuous rotation. The air between the spinning heat sink and the contact plate will conduct magnitudes better than standing trapped air. Whether or not it's enough to compete with our idea of high end cooling remains to be tested.

[Kain665]

Create an impractical heat sink they said...




It would be fun, they said...

[STEvil]

What remains to be seen is whether the conductivity is increased only to the air bearing and out through the air movement or if the air bearing also conducts to the rotating portion of the heatsink as well.

If it does conduct to the rotating mass of the heatsink as well it may be feasible to create a heatsink with a vapor-chamber like base with just a spinning disc in it to create the air bearing effect thus possibly greatly increasing the effective heat energy conducting surface area... maybe.

[SubZero.it]

He does say in the video that the high pitched noise is only because its got an open motor while they are testing - he then turns the motor off to demonstrate how quite it will be when the motor design / cover is finished.

I watched the entire video and I saw that part. It was still loud.

[Kallenator]

One does not simply judge fan noise on video of a prototype on Youtube...

[STEvil]

I watched the entire video and I saw that part. It was still loud.

You should turn your sound down from 100% volume.

[[XC] Lead Head]

Indeed. I had to turn my volume up the point where the narrator's voice was uncomfortably loud before it sounded anywhere near "loud"

[NEOAethyr]

Supposedly they made these things to compete with some of the bigger heatsinks, for smaller applications such as leds.
I wanna see the temps of it on a 100w led...

Edit:
Gonna some vids about it instead of reading on it.

[Solus Corvus]

What remains to be seen is whether the conductivity is increased only to the air bearing and out through the air movement or if the air bearing also conducts to the rotating portion of the heatsink as well.

The extreme shear between the base and top causes a sort of forced convection that transfers air molecules between the two. It's talked about in the whitepaper.

If it does conduct to the rotating mass of the heatsink as well it may be feasible to create a heatsink with a vapor-chamber like base with just a spinning disc in it to create the air bearing effect thus possibly greatly increasing the effective heat energy conducting surface area... maybe.

FINALLY! Someone really gets it. If the numbers are true then this cooler is about the size of a stock Intel cooler while performing on the level of tower coolers. So what happens when this tech gets machined out of copper instead of Al, gets heatpipes or a vapor chambers, or the concept is applied to water?

As for the other points:

1. Vertical mounting:
Yes, it will work. Air bearings are extremely stable and very effective at maintaining a gap. The spindle acts as a height limiting stop and a fulcrum that the top can tilt on. But any tilting of the top creates a rise in pressure on one side of the gap and a corresponding drop in pressure on the other side. Air bearings are used in hard drives and they have no problems being mounted vertically.

2. Sound level and power consumption:
First, as mentioned in the video there isn't a cover on the motor so you can hear it whine. But there are other factors too. The prototype is using a commodity motor, while fans usually use custom windings. During the process of development the exact number of windings on the motor, the number of fins on the top, the size of the airgap, etc still has to be optimized.

3. Conspiracy theories:
All I can say is: LOL!

4. Dust:
It won't be dust free. But it would have way less dust than a stationary heatsink. Fans are usually relatively clean compared to the heatsink underneath.

5. Personal thoughts:
A major potential weakness is the air ports that pressurize the air bearings. If the passages have filters in them then they could clog and the air bearing would lose pressure over time. If there are no filters then dust particles will find their way into the airgap and it will have to be periodically cleaned.

But mainly, if the numbers are right for air, then I really want to see what the same concept could achieve with water.

[STEvil]

Yes, but it remains to be seen if the shear/bearing has air exchange to a significant degree, or if it is only between the finned area and the energy point, is what I was trying to get at.. hence my second part about the vapor chamber-like possibility.

Another thing to look into would be texturing the base of the rotating portion of the heatsink or the disc in the shear/air bearing chamber. Maybe a pattern like a golf ball would be best, maybe holes, maybe like a cheese grater? hm..

[Solus Corvus]

If it works at all it is because heat is being transported from the base to the top. The whitepaper talks about how the shear between the base and the top causes convection between the two. The volume of air flowing through the gap wouldn't be enough to cool anything.

[SubZero.it]

You should turn your sound down from 100% volume.

This doesn't make sense. Turning down volume doesn't make something quiet.

[Liam_G]

yes, but having volume set at 100% doesn't necessarily reflect how loud something is in real life. if you have your volume at 100% and normal talking voice sounds like yelling don't try and tell me that the whisper quiet motor is now too loud. I watched the video and the motor noise was audible and annoying, but it will be shielded and not be so bad when it is finished, there will still be some noise, however the spinning heatsink itself is virtually inaudible when the motor isn't running, or when it is properly shielded.

[SubZero.it]

The answer is two posts above.

[CrazyNutz]

At 3:45 you can hear how loud it is in respect to his voice, so volume does not matter. Unless his voice was dubbed in, or they were using high compression in which case you would hear
a lot of background noise.

If it does cool better than say a stock intel HSF but not as good as higher-end aftermarket, what is the market for this? 1u servers? I think not.
You would need to have gaps between the servers to allow air to enter the top, rendering it impossible to utilize this in high density server farms

This may work better for GPU, however I suspect hanging this thing upside down will cause an issue with the air bearing.

[Manicdan]

i think in servers you will just have a problem with how air flows through everything. a tower cooler for lateral air is going to be the best way to circulate cold to hot. this thing with vertical to lateral airflow will probably just overheat unless things are spread out enough (meaning only works well in open benches)

[jogshy]

Ok! Some time passed.... Where are the real Sandia-based-licensed coolers from Thermalright, CoolerMaster, etc? I need one for my upcoming Trinity APU HTPC

[CrazyNutz]

Ok! Some time passed.... Where are the real Sandia-based-licensed coolers from Thermalright, CoolerMaster, etc? I need one for my upcoming Trinity APU HTPC

Precisely!

[Greg83]

3 months doesn't result in a consumer product from a functional shown prototype.
not even when its within the same company.
if proven effective, it will first have to be sold to intel amd for high end desktop processor and server markets.
then with mass production to cover licencing costs for the r&d. it will find its way down into the after market coolers in 1-2 years.

[jogshy]

then with mass production to cover licencing costs for the r&d. it will find its way down into the after market coolers in 1-2 years.

In 1-2 years the desktops won't even need a cooler because ARM (or graphene) gonna take over the world !
I want my Sandia noooooooooow !

[Ivanlabrie]

Sandia in spanish...

[kuroikenshi]

Now that's one way to keep cool, with "Rodajas of Sandia".....OMFG!!! someone pls lock this before people begin posting!!!

[STEvil]

I dont expect we'll see this in consumer applications due to filtration requirements for the air.

[bhavv]

Even if it works, I dont think it will be as good as a current high end CPU cooler, or a self contained water cooler.

And it looks far too dangerous and risky, I'd rather go full water cooling before thinking of using one of these. It would cut through any cables that accidentally land on it too fast, if the blade comes loose loads of stuff in the PC will die, and also in the video it only demonstrates it working upright, its going to be a lot trickier getting it work sideways for tower cases.

[El Mano]

So Sabertooth's plastic cover finally has a function