[Gllug] Re: Super Computing Time

[John Hearns]

dennis furey wrote:

>A clock frequency of 560 GHz implies a period of 1.78E-12 seconds,
>during which a clock signal propagating at the speed of light (2.998E8
>m/s) covers about half a millimeter. Anyone who can squeeze all that
>gear into a space smaller than that might be a pretty good physicist
>already. ;)
>
I see the smiley,
and I'm sure that you got the point - AFAIK this is a cluster,
and all that's been done is to sum the clock speeds of the indivudual
processors. Probably, and I don't like to start flinging insults at all, not
a very helpful measurement.

By the way, talking of physics, a very good rule of thumb
is that light travels one foot per nanosecond.
(one foot is about a third of a metre).
So if you are setting up cables to do detector readouts this
can come in handy.



-- 
Gllug mailing list  -  Gllug at linux.co.uk
http://list.ftech.net/mailman/listinfo/gllug