jcoates at gmail.com
Thu Oct 23 18:32:24 MDT 2008
>I have to ask, though: have you seen a government datacenter?
>What about one that spanned acres? :)
i have. the 'compound' was acres in size, but the actual floor space was
measured in 1000's of square feet, just like a commercial data center.
really amazing super computers and storage facilities - but to be honest,
the most impressive thing about the whole place was that the entire grounds
of the facility was draped with a net.
that's right - imagine a small business park that spanned 10 acres or so,
and then imagine the entire place with a net 'ceiling' about 40 feet in the
air. i asked them what was up with the net and they said it's to stop
people from infiltration from the air. i guess that would work.
but i guess my original question about the 95% guesstimate relates to the
absurdity of that particular way of measuring things in this context. in
terms of computational power, the difference of 5% is meaningless (not to
mention a completely arbitrary, made up number.) eg. 95 THz of processing
power vs 100 THz of processing power is not a significant difference when
applied to problem solving.
so, unless you want to blow the credibility of your well crafted views on
civilian computational power vs. the governments power, i suggest you come
up with a difference way of expressing this value.
but yeah, after seeing several government labs, i tend to agree that
generally speaking, the governmental computational power is *not
significantly* disimilar to commercially available power. but then again, i
don't really know, it's just my guess based on what i've seen at government
labs and my low opinion of the ability of our government to innovate
*independently* of the private sector.
More information about the PLUG