Standford study on arrays for direct die cooling

Standford study on jet arrays for direct die cooling

http://www.nnf.cornell.edu/nnun/2002NNUNpg28.pdf

Too crazy for me but some may enjoy

I will try my best to put it in layman’s terms. From what I gathered they are working on a multiple channel design for the CPU waterblock. In current block designs there is only one channel for the water to flow though. This design would put multiple channels inside the block. In the case of the Stanford project they created a block with 40 channels for the water to flow through. Think of it this way. An aircooled heatsink with 40 fins does an immensely better job of cooling a CPU than a heatsink with only 1 fin.

Those Stanford guys don't get out much do they. They're just figuring out channels, in a few years it'll be impingment...

Somebody really should email them some links.

EDIT: Gah! I really should have read the paper before I commented. They are working on jet impingement

Volenti and Jaydee116 were doing jet array impingement for direct-die well over a year ago. Their work was part of the inspiration for the Cascade.

However, what these Stanford guys are doing is micro-jet, as in jets of well less than 0.1mm in size. They are trying to achieve adequate cooling using super-low flow rates (0.03GPH - yes gallons per hour). A fairly different scenario to be sure.

Why are they limited to that flow rate? Is it even possible to get sub-100 degree temps with that?