QuoteResearchers from Sandia National Laboratories have created what they describe as "a radically new architecture for air-cooled heat exchangers", incredibly efficient compared to the traditional fan+heatsink setup, quiet and even immune to dust.
Sandia Cooler achieved such a feat with a quite simple design where the metal heatsink itself is the fan: "heat is efficiently transferred from a stationary base plate to a rotating (counterclockwise) structure that combines the functionality of cooling fins with a centrifugal impeller", Sandia Labs official site explains, while the cooler is powered by a "brushless" motor in the middle of the device.
The new cooler is 30 times more efficient than a traditional setup, Sandia Labs researchers state, 10 times smaller compared to a high-end commercial unit available nowadays but with the same cooling performance. Furthermore Sandia Cooler is immune to dust layering because it spin at a constant 2000 RPM speed, while the centrifugal forces constantly push the dust particles out of the device.
Among the potential applications for its cooler, Sandia Labs – a wholly owned subsidiary of Lockheed Martin Corporation and two major research and development laboratories belonging to the United States Department of Energy – list laptop and "high performance 'gaming' PCs", gaming consoles, LED lightning, automotive, "other" electronic devices and "any device comprising one or more forced-air exchangers".
The market debut of Sandia Cooler should happen soon, too: the researchers say that two yet-to-be-named companies – one working in the computer cooling business and the other one in the LED lighting business – have already licensed their promising technology.
Wouldn't want to stick your finger in it :P
Source (http://www.neowin.net/news/next-generation-heatsink-has-no-fan-at-all?utm_source=twitterfeed&utm_medium=twitter)
but...
how do you get the heat off your chip and onto the heatsink to start with ?
or did I miss that bit ? flicked through it twice, but cba to watch it all again....
if it's floating on a bed of air... that'll make it even worse ?
They talked about it but didn't quite get it myself. You can see a block with 4x heatpipes but don't know where they go and how they distribute the heat to the moving sink and as you said I cba to watch it again :P
Its says the heat is transfer through an 'Air Bearing' I think, he was talking about air and heat dynamics and things like that so the science must be there, its and interesting idea and will be interesting to see some real world results.
Can just imagine it coming loose while spinning and taking out your gfx card :p
Sent from my SK17i using Tapatalk 2
It seems to me as though the act of the impeller lifting off pulls the heat with it somehow (I think because of the change in pressure where it draws the air up through the inner ring and outwards through the fins, the pressure change must transfer the heat with it). Damned if I know for sure though, those guys are brainboxes. From reading their documentation it seems you can mount this in any orientation you like, so it doesn't rely on being horizontal either.
So, who else is investing in their shares? :lol:
but... even under ideal conditions... heat transfer from the chip to base to the air to the cooler will be crap compared to from a chip to a normal heat sink
Apparently not, if their documentation is accurate. This explains it a bit better I think:
http://ip.sandia.gov/technology.do/techID=66 (http://ip.sandia.gov/technology.do/techID=66)
there's a good diagram on the fact sheet doc too: https://ip.sandia.gov/techpdfs/Sandia%20Cooler-Fact%20Sheet%20for%20the%20Microelectronics%20Market.pdf (https://ip.sandia.gov/techpdfs/Sandia%20Cooler-Fact%20Sheet%20for%20the%20Microelectronics%20Market.pdf)
I kind of get it but its all science to me! :P
Basically they say that even with the trapped air the thing is better than the standard setup. Once a couple PC hardware sites get their hands on them and do some OC testing we will know if its fail or not and go from there :)
Thing is this isn't exactly some BS no-name company that might otherwise make you instantly think of snake-oil, the fact that its a pretty big subsidiary of Lockheed Martin suggests it's probably going to do exactly what it says on the tin. I hope so anyway, as nice as my Noctua is it's a bugger for size and awkward to fit, this kind of cooler is a great step forward, I'd pay £100 for one of these if it had the potential to be bracketed for different socket types (i.e future-proofing).
Brackets are pretty standard these days so shouldn't be a problem and as they are taking orders for AC and other companies it shouldn't be too long before they appear on the PC market.
interesting, from a cursory glance it looks like they're arguing that the thermal resistance of the "air bearing" is more than compensated for by the reduction in boundary layer thickness around the fins. Sounds plausible.