Purdue Develops “Intrachip” Micro-Cooling System

Megalith

24-bit/48kHz
Joined
Aug 20, 2006
Messages
13,000
Conventional chip-cooling methods use finned metal plates called heat sinks, which are attached to computer chips to dissipate heat. Such attachment methods do not remove heat efficiently enough for an emerging class of high-performance electronics, however, so researchers have developed a new type of cooling system that circulates a liquid coolant directly into electronic chips through an intricate series of tiny microchannels.

“You can pack only so much computing power into a single chip, so stacking chips on top of each other is one way of increasing performance,” said Justin A. Weibel, a research associate professor and co-investigator on the project. “This presents a cooling challenge because if you have layers of many chips, normally each one of these would have its own system attached on top of it to draw out heat. As soon as you have even two chips stacked on top of each other, the bottom one has to operate with significantly less power because it can’t be cooled directly.” The solution is to create a cooling system that is embedded within the stack of chips.
 
Great until those microchannels are blocked..

While I agree, they could always use something akin to a heatpipe or vapor chamber setup. Other than developing a leak I have never seen one fail.

After briefly looking at the article, that is exactly what they are doing.

Here is a nice little snippet from the article... crazy!

A fundamental requirement stipulated by DARPA is the ability to handle chips generating a kilowatt of heat per square centimeter, more than 10 times greater than in conventional high-performance computers.
 
  • Like
Reactions: Azrak
like this
Welcome to 5 years ago when this was first proposed with lab prototypes.

There was a problem with the design however. It required an ultra pure cooling liquid and over time all liquids eventually decay and contaminate. That and the micro channels between layers increased the inter transistor distance which is a big deal.
 
I can see what they are trying.
I think it is similar to how we build high rises to pack more people into smaller footprints, except this is for chips.

I don't think this scales all that well, because the more chips you stack, the more channels you will need to cool the chips that sit further below. This also means that chips that sit further up will have to be even larger to accommodate the extra holes, which defeats the purpose of stacking the chip to reduce footprint.
I am not a chip designer, but don't longer traces reduce the potential IPC of the chip?
A chip with lots of perforations sounds like it will have longer traces than a normal design.

So, maybe not that important for high power processors, but it sounds promising for lower power processors, DRAM and flash.
 
Great until those microchannels are blocked..

New tech has to start somewhere.

If anything, the liquid should be sealed in the CPU package, then another liquid to liquid heat exchanger external to the die would transfer the heat to your 3rd party CPU cooler... or something like that.
 
Back
Top