Re: [TCLUG:19672] VALinux vs Compaq, and OT: statistical inference

[Philip C Mendelsohn <mend0070@tc.umn.edu>]

On Wed, 19 Jul 2000, Carl Wilhelm Soderstrom wrote:

> > Not entirely.  The crossbar switch was something developed by DEC, and
> > I'm pretty sure that Compaq just inherited it in the purchase.  But I
> > don't doubt that the same performance is achieved elsewhere, whether by
> > the same means or not.
> 	I belive my 1984 copy of Kai Hwang's "Computer Architecture and
> Parallel Processing" mentions crossbar switches, in the contex of the
> best (and most expensive) way to build a system bus. I don't recall it
> being particular to any manufacturer.

I'll check my facts.  I know that there was *a* crossbar switch developed
by DEC, and it was a big huge project.  I think it earned a couple of
patents, but no doubt there is more than one way to skin a cat with that
many variables.

-- 
Lottery:    a tax on people who are bad at math