[Info-vax] Out with Hurd, in with OpenVMS

[Michael Kraemer]

In article <t21JfrxoQRgS at eisner.encompasserve.org>,
koehler at eisner.nospam.encompasserve.org (Bob Koehler) writes:
> In article <i4g6uh$434$03$1 at news.t-online.com>, Michael Kraemer
> <M.Kraemer at gsi.de> writes:
> 
> > Istr discussions even among DEC/VAX bigots why in the world
> > one would spend lots of money on the likes of VAX 6000/9000
> > when a VS 4000 could do the same - at a much lower price
> > and you could put one on each desk.
> 
>    You got a VS 4000 with a vector processor and the I/O bandwidth
>    of a 9000?
> 
>    Mind you, I had no trouble understanding when a nieghboring site
>    went from it's single CPU 9000 with a vector processor to a few 
>    single and double CPU 6000 with vector processors.

And what if people's codes aren't vectorizable at all
(as is the case for most codes out there)?
And if having a bunch of moderately powerful graphics-enabled
machines is better bang for the buck than a single big iron?