Sandia Air-Bearing Heat Exchanger

[Duper]

This looks like it's going to be pretty cool. (no pun intended..but what the heck )
It's looking to replace the standard CPU heat sink.

CLIK!

This is a link to Atomic PC. There are links in the article for the actual white paper on this.

[Grendel]

Interesting. But I'm not holding my breath, I have seen one too may cooling solutions claiming to "obsolete the classic heatsink" in my lifetime... As for this one I doubt it'll replace the traditional FFHS solution anytime soon, the manufaturing costs will make it quite expensive due to the strict tolerances required to make it work.

[snoopy]

I'm sceptical. The challenge for this design is to have a bearing that has both low friction and low thermal resistance. I don't think it's as easy as they want you to think it is. Air is naturally a really bad thermal conductor, so it would seem to me that an air bearing would present a decent heat barrier, regardless of its thickness.

My 2c: if you want to replace FFHS with something more efficient, liquid cooling is that way to go.

[Foil]

My 2c: if you want to replace FFHS with something more efficient, liquid cooling is that way to go.

x2.

[Tunnelcat]

Phase change would be better, but right now, it's too big of a setup for the standard computer case and has condensation issues. What we really need is to come up with a new more efficient way to do computing, without generating so much heat we can fry an egg.

What if in the future every home had a central built-in computer that has it's own cooling system? The computer could then be accessed by remote terminals in the house. A single central system would be easier to keep cool, clean and maintain, just like a furnace or air conditioner and you wouldn't have a giant wart sitting on your desk either. As a bonus, in the winter, the waste heat from the processor could be put to use heating the home and in the summer, central air could be used to cool the processor. Just a thought.

[snoopy]

Phase change would be better, but right now, it's too big of a setup for the standard computer case and has condensation issues. What we really need is to come up with a new more efficient way to do computing, without generating so much heat we can fry an egg.

What if in the future every home had a central built-in computer that has it's own cooling system? The computer could then be accessed by remote terminals in the house. A single central system would be easier to keep cool, clean and maintain, just like a furnace or air conditioner and you wouldn't have a giant wart sitting on your desk either. As a bonus, in the winter, the waste heat from the processor could be put to use heating the home and in the summer, central air could be used to cool the processor. Just a thought.

Cool idea for one of us. For the masses, I see cloud computing catching on a whole lot earlier.

[fliptw]

Here's the thing: the air bearing isn't static. The bearing only exists when air is moving.

The air hockey table is an apt analogy.

[Sirius]

What if in the future every home had a central built-in computer that has it's own cooling system? The computer could then be accessed by remote terminals in the house.

If you really felt like it you could pretty much already do this kind of thing with this, though I'm not sure it's quite ready for home users yet. With the way Google and MS are going we're much more likely to see the cloud provide the central hub and thin clients/chromebooks/whatever the access points. Home servers are great for power users, but most consumers aren't power users, so I'm not sure how far that'll go.

[Krom]

My 2c: if you want to replace FFHS with something more efficient, liquid cooling is that way to go.

x2.

Water coolers don't inherently perform any better than air coolers, the reason they are usually more effective than air is because by actively pumping the water they can hide its relatively poor thermal conductivity and exploit its excellent thermal capacity to scale up to absurdly gargantuan surface areas on the radiator without loosing any significant efficiency.

Traditional heat sinks are limited by how well they can transport heat from the source to the extremes of the fins, since they ultimately depend on just passive thermal conductivity they can only transport the heat so far away (a few inches is the limit). Water coolers avoid that limitation since they literally actively pump the heat away from the source which lets them maintain the same efficiency even transporting the heat considerable distances.

It all comes down to surface area, most water coolers simply have more of it than traditional heat sinks.

[Tunnelcat]

What if in the future every home had a central built-in computer that has it's own cooling system? The computer could then be accessed by remote terminals in the house.

If you really felt like it you could pretty much already do this kind of thing with this, though I'm not sure it's quite ready for home users yet. With the way Google and MS are going we're much more likely to see the cloud provide the central hub and thin clients/chromebooks/whatever the access points. Home servers are great for power users, but most consumers aren't power users, so I'm not sure how far that'll go.

Well, the cloud is fine if you have a good, reliable high speed internet connection. But living out in the sticks, as far as internet speed goes, it's either Century Link and the1.5 Mb/s I have now have, or Comcast, which everyone shares in my neighborhood, which bogs down when everyone uses the net. If they're forcing us into using the net for everything, at least install good high speed fiber pipes or 4G WiFi to access it.

I still think in the future that computers will become some type of necessary built-in home appliance that runs things as well as communicates with the net and allows for entertainment access. But, like you said, it's complex and not ready for the average user yet.

[TechPro]

It's true that it's an interesting idea, but (IMO) it has a couple obstacles it would have to overcome before it could be practical with computers.

1. Gravity. It can only work well if the object to cool is under it horizontally (not vertical, not upside down, not on an angle).

2. Requires space. Not small enough for laptops or very compact locations.

3. If an air gap even with rapidly moving air was sufficient to cool most processors without direct contact with a heat sink ... why have we spent so much time with thermal compounds and tightly fastened heavy heatsinks? Because moving air is often NOT nearly enough to cool a lot of processors.

Interesting science, but I don't think we're going to see it much on computer stuff. Besides, many new devices (like the iPad or HP's Touchpad) do not require methods like that to cool them. In fact the newer portable processors are more and more energy efficient (thus longer battery life) which means less heat to disperse anyway.

Just my opinion.

[ Post made via iPad ]