MSI's Stirling Cycle CPU cooler....

[Dr Mordrid]

http://www.techspot.com/news/29214-msi-motherboard-transforms-chipset-heat-into-fan-power.html

MSI motherboard transforms chipset heat into fan power

MSI’s main theme for this year’s CeBIT is “ECOlution”, short for eco-friendly and environmental evolution. The company will be showcasing their next-generation Intel 4 series motherboards, a green concept notebook designed with Anion technology, a gaming series notebook with Turbo Battery Energy-Saving Technology and a selection of low-cost PCs.

But what MSI hopes will really catch the eye is its new electricity-less motherboard fan concept, which is powered only by heat and air. Dubbed the “Air Power Cooler,” this unique prototype operates on the Stirling Engine Theory to transform the thermal output of a chipset into the kinetic energy necessary to power that same chipset’s fan and consequently cooling its heatsink.

MSI’s air power cooler will reportedly transfer over 70% heat power into air power. While this is all good news on the green-front, it’ll be interesting to test this concept with hotter components such as a CPU. MSI hopes to turn this early concept into mass production in the near future.
>

[cjolley]

You know, even if that really doesn't work I like it just for the idea.

[Claymonkey]

You know, even if that really doesn't work I like it just for the idea.

Agreed! Though that system looks like it would be best for cooling an underclocked Pentium-m not a C2D or an X2. Still....

[VJ]

Actually, it is currently considered as a cooling for ... the northbridge rather than for the CPU...
(and the nortbridge is is passively cooled on many mainboards...).

Don't get me wrong, I really like the idea, and it would be great if they could adopt this technology for CPU's and GPU's, but it raises some questions:
1. how will dust affect the piston?
2. what is the lifetime of the piston?
3. what is the noise produced? (it will not be the sound of an electrical motor, but of a piston moving...)


Jörg

[Lambo-Fan]

On the one hand the concept looks really innovative to me. On the other hand I also see the problem that dust probably will stop the airflow easily. Futhermore I guess that a contruction like this one always needs a certain size to work properly - which means that it would be very difficult to use it on a graphics card (let alone the fact that it wouldn't create enough airflow to cool a modern GPU or CPU anyways).

the other Jörg

[Evildead666]

I don't see it cooling a cpu any time yet.

As i gather, a gas is heated, which expands, and pushes a large piston up.
Then the gas cools and contracts, pulling the piston down.
repeat.

That would mean that the nb chip would have to heat the gas up before turning the fan, so there would be less fanspeed on a cool chip, and more fanspeed on a hot chip.

i still can't see how letting a chip get hot to cool it is going to work...

[Dr Mordrid]

What temp the required phase change occurs at depends on the material. Some do go through it at high temps, but others do not.