Bell Labs Plants Nanograss to Cool Mobile Chips

LoadWB writes "TechWeb has an article about Bell Labs' new liquid cooling technology for mobile processors. The tech, called 'nanograss' is described as 'tiny tubes that spray liquid on chip hot spots.' The use of this cooling technology reduces the power required to actively remove heat from mobile processors. Other applications are possible, but it seems it was primarily developed for use with mobile CPUs."

Comments

Maybe Apple can use this

[badriram]

Apple could maybe use this and come out with those wonderful g5 powerbooks some time soon.

Ahhhh.... G5 powerbook....drooling...

Hmm. . . .

[ookabooka]

Brings a whole new meaning to watering the grass. . .
Seriously though, its nice to see some new heat dissipation technologies. . . but it still comes down to how much thermal energy the chip pumps out. . . this is merely equivalent to a more efficient fan/heatsink. Though it should keep the chips at a cooler temperature (compared to their standard air counterparts) your laptop is still gonna get way too hot to put on your lap.

Re:Hmm. . . .

[moreati]

If it works, it will push the envolope, improving not just total power consumption, but weight, volume & temperature also.

We already have water cooling at the macro level - a radiator + pump + heat exchanger + resevoir system will give a lower temp and less noise for the same or better heat removal capacity eg . these.

The improvement this would provide is watercooling at the micro level, just to the most critical components. The improvement in heat conductivity from the chip to the cooler should mean lower temps for the same transfer. Cooler.

The bore of the tubes implies 50 ml liquid, rather than upto 1 litre (2 pints) currently used. Lighter.

Less water for the same heat transfer means a smaller pump. Lower resistance in the chip due to lower temps would mean less power disipation. Longer running on batteries.

On the air side (dissipation from the cooler to the environment), heat exchanger tubing with ~100 micron diameter (the artivle soesn't say they've done this, but it seems a logical extension) gives enourmous surface area/unit volume, giving better dissipation for the same airflow. Quieter.

So I would surmise this is ideal for laptops, it improves all 3 of the key features - weight, longevity and actually-able-to-use-it-on-my-lapiness.

Re:New error message....

[prat393]

Pretty soon there'll be a gasoline powered chip, or at the very least a laptop fuel cell.

Technology: applied ingenuity in order to keep everything exactly the same.

Will the pipes clog?

[Anonymous Coward]

It's cool, but it sounds awfully complex. Wouldn't small amounts of impurity gum up the works pretty quickly?

Wow

[Un0r1g1nal]

This looks like really good stuff, being able to localise the temperature dissipation would be handy for lots of technologies. I hope that this one gets developed fully and hits the markets soon. The better the cooling capacity the more we can clock our chips :)

As for having to refill the cooling agent periodically, I doubt that this would be a problem with mobile phones, this would be a completly self contained cooling system, much like a heatsink is today, (only a heatsink doesnt have a liquid running around the inside of it :P). The likelyhood is that by the time the liquid would need replacing - if ever, the phone would be at lesat a few years old, and so the owner would probably have it lying around in some drawer since they got their brand spanking new top of the range all singing all dancing holographic video phone...

Re:Where does the energy go?

[Awptimus Prime]

but it doesn't really seem to be such a great innovation as it is being hyped up to be..

Thanks, you saved me from starting a thread on that issue. I kept thinking "well, where does all this fluid go to cool off?".. The amount of fluid and circulation will be dependant on the CPU. If it's a hot running P4 or AMD64, then the resivor needed for proper cooling will be similar to that of a normal water cooling kit. While eliminating the need for a water pump, which can be made tiny enough to be insignificant compared to whatever you have to use as a radiator to dissipate the gathered heat.

Personally, I'd trust a water cooling rig before something like this. Due to the teenie, tiny nature of the tubes, I could imagine any microscopic particles in the fluid would eventually clog it up. For instance, the chemicals released as the processor ages would be likely to collect and clog an area a few microns across, easily. Since it's a passive system, there would be no means to flush the blockage out via the pump. At least with normal liquid cooling, the user can repair problems before they cripple the system. With a solid-state solution like this, you'd be dead in the water.

Re:What's next, the nano-bong?

[Rostin]

I'm confused, because it sounds like the "nanograss" is some kind of little temperature-actuated valve. More likely: The real idea is not that liquid is applied only to problem areas, but that heat-removal is applied only to problem areas. Channels lined with tiny fins are cut all over the die, and liquid is pumped all through them. Liquid flowing through channels in cool areas is heated less than liquid flowing through channels in hot areas. The energy savings comes from needing a lot less coolant (meaning less cooling and pumping), which is in turn a result of the intimate contact between the coolant and the die, and the huge heat transfer area from all those fins (relative to what you'd have w/ just empty channels).