Intel Cooling

JJKizak wrote on 6/6/2008, 5:48 AM

After reading in one of these posts about Intel creating tiny nanatubes to run water through to cool the new processors I have really wondered why they just don't print in layers nano thermocouples and reverse the applied voltage like the Appollo spacecraft did about 50 years ago? Borg/Warner has been selling these kinds of cooling units ever since for cooling electronic cabinets.
JJK

Comments

apit34356 wrote on 6/6/2008, 6:01 AM

IBM has a working model of a 3D ics towers being h2o cooled with micro channels. The IBM design permits ics to be stacked and look like a city inside one cooled "packaged" ic. One problem with nanotubes and H2O is that H2O becomes HO +H more often than previous thought ( but if thinks out the acid+H2O relationship, it becomes more obvious) , permitting HO & H to interact with the carbon or worst, passing H thru the nanowall.

blink3times wrote on 6/6/2008, 6:18 AM

I'm on liquid cooling now, but when I first started looking at the different cooling options (ie: peltier cooling, fans....) it became clear that liquid cooling is far more efficient in terms of price, complication, versatility, and overall ability.

With liquid cooling you can completely remove the heat at the source WITHOUT having it float around inside the casing warming everything else up before it exits. This makes a HUGE difference in the overall running temp of the HDD, video card.... etc. It's also cheap. I built my cooling system from scratch for far less than what a large fan and heat sink would cost.

I do however wonder how easy it would be to physically plug a nano tube with scale, impurities.... yadda, yadda.

apit34356 wrote on 6/6/2008, 10:01 AM

trying to get the heat of a multi-ic out is a serious issue and nanotubes are a hot topic in research and design circles. But one protein or crystal, the tube is dead. but....... checking out possibilities is a good thing! ;-)