Lawrence Livermore Labs used to be heated by the output of the computers
they had. At one time, I believe, in the 60s and up to I don't know when
they had at least one of every mainframe and supercomputer around. Lots and
lots of heat.

Or so I was told by folks who worked there and the Berkeley Lawrence Labs.

Mike

On 4/16/10 3:20 PM, "Peter DiFalco" <peter.difalco@GMAIL.COM> wrote:

> http://www.livescience.com/technology/Water-Cooled-Supercomputers-Internet-100
> 415.html
>
> FTA: "Since water and electronics don't mix, the coolant water in Aquasar
> does not ever directly touch the silicon computer chips themselves. If a
> leak should occur, internal sensors would shut the machine down before
> causing a short circuit."
>
> Hey, I wonder why they don't just let it set off a buzzer and make white
> smoke...
>
> "The tubes interspersed in Aquasar will link back to the primary water
> transportation network where some of the collected heat will then be
> passively released back into the heating system of ETH Zurich."
>
> ....Which implies that the building's occupants will need to wear their
> earmuffs and overcoats until the programmers have had a chance to drive the
> supercomputer around the block for half an hour.
>
> The real question is whether they'll still have toasty feet even while the
> air conditioning's on in the summer.
>
> Happy Friday!