🚀 go-pugleaf

In article <856@cadlab.sublink.ORG> martelli@cadlab.sublink.ORG (Alex Martelli) writes:
>A definition I remember, and which seems to me to make sense: a supercomputer
>is a system architected and designed for PERFORMANCE, to the *detriment* of
>PRICE/performance.  The curve you can draw with price on X axis and
>performance on y axis will generally have an inflection point: below that,
>you are below optimal price/performance because you are not exploiting
>economies of scale, and above that, you are below optimal price/performance
>because of diminishing-returns effects.  
>I always like this idea, incomplete as it is, because to me it says something
>about the ('politically' motivated) "wisdom" of buying a supercomputer, and
>then overloading it with timesharing jobs...

Not bad, Alex.

I recall sending Eugene a blurb on the subject more than a few months
ago - still got it, Eugene?  I don't...

Anyway, a supercomputer...

...should cost more than a million dollars (covers lots of mktg overhead)

...should come with a small armada of analysts/CEs/salestypes (kinda
   like a Naval carrier group)

...should go on PM at least once a week (aka CE job security:-))

...should always be overloaded

...should always have at least one session running "vi" on it

...almost always makes for great plumbing/air handling architectures

...and most of all, is always but always surrounded by politics,
   money machinations, multi-level RFQ/RFP/RFI/RFwhatevers, assorted
   sales/marketing bonanzas, and other generally fun and unique
   intrigue and/or occasional international espionage/scandal.

You say you want a technical definition?


-- 
Rob Peglar               Network Systems Corporation
Internetwork Group       7600 Boone Avenue North
robp@anubis.network.com  Minneapolis MN 55428   (612)424-4888 x1028