View Single Post
Unread 11-25-2003, 12:36 PM   #10
Starman97
Cooling Neophyte
 
Join Date: Oct 2003
Location: Austin,TX
Posts: 27
Default

The real problem here is going to be..
How do you get almost 200Amps
into the part without melting some pins or trace. At 200A, even just a little resistance is going to get hot, fast! A poor connection is going to go incandescent as it heats up and the resistance increases, causing more heat...

90nm parts are going to sub 1V Vcore's,
so every Watt is more than an Amp of current.
Designers are going to have to think about cooling the Vcore supply, and the power connections.

This reminds me of something I read in one of Drexler's nanotech book about future super computer CPU cores being a 3d chip, maybe a foot on a side, two sides have copper cables, carrying as much as a megawatt of power, one side is a high pressure water inlet, the other side, live steam outlet 1 side for fiber-optic IO and one side to screw it into a support structure.
We're getting there..
Starman97 is offline   Reply With Quote