rack power question, and a prediction about "direct heat removal" (DHR)

Marshall Eubanks tme at multicasttech.com
Fri Apr 4 12:58:52 UTC 2008

On Apr 4, 2008, at 5:14 AM, <michael.dillon at bt.com> wrote:
>> My guess is that someone will come up with an inexpensive,
>> reliable way to put a heat collector, which will basically
>> look like a car radiator the size of a rear rack door,
>> directly behind the hot air coming from the systems in the rack.
>> b) air is a lousy way to transfer heat away
>> from where you want it to be compared to other materials.
> So how did you propose to get the heat from the equipment
> to the car-radiator door?
> The idea of making the entire room into a refrigerator by
> bringing coolant to the racks is a good one, but I think
> that for this to be successful, more attention needs to be
> paid to physical placement of things, and the chillers need
> to be broken open. By that I mean that chillers cease to be
> a big box at the edge of the room because they are now part
> of the room itself. Think of a flat chiller attached to the
> ceiling with spaces for racks to be inserted into it.

Dealing with heat transport is an important part of the design of  
spaceflight hardware, with no
convenient gases to extract heat in most cases.  Heat pipes, which  
can have thermal conductivities much
higher than solid copper, are commonly used to transport heat to
outside radiators, but more important is that the entire system is  
engineered to remove the heat of operation from where it is created  
to where it can be radiated away.

It sounds to me that blade systems are reaching the point where heat  
transport will also have to be designed into the rack system, and  
maybe also into the room system. A set of heat pipes could, for  
example, remove heat from blades and bring it to the floor of the  
rack, say to a radiator directly into the plenum of the airflow.

The working fluid inside heat pipes is unlikely to leak, and  
typically contains very little fluid, but in any case could be chosen  
to be something that would be fairly benign if the pipe was breached,  
say methanol.

There might be a business model here...


> --Michael Dillon

More information about the NANOG mailing list