Re: Good news for SPARC/Linux

David Feuer (feuer@his.com)
Sat, 12 Dec 1998 21:38:07 -0500

ralf@uni-koblenz.de wrote:
>
> On Fri, Dec 11, 1998 at 01:59:16PM -0500, David F. Newman wrote:
>
> > > Needless to say, I want one of these... though I wonder if there is _any_
> > > real-world application which 14 CPUs would speed up 14 times.
> > >
> >
> > I could easily use all 14 CPUs when ray tracing. Either having
> > each CPU render an individual frame of an animation, or giving
> > each CPU a piece of a still image to render. It is highly
> > math intensive with io only when an image is completed. A
> > perfect application for SMP.
>
> This is almost the perfect application for clusters. Clusters don't have
> the tight communcation network of SMPs or ccNUMA but tend to be quite
> cost efficient which is a good fit for the typical raytracing app.
>
> Ralf
>

Clusters also don't suffer from the bus clogs of shared-memory machines.

I'm curious: are there any (cheap) systems out there with some shared
and some unshared memory, and multiple buses? This would seem to allow
a wonderful combo of the speed of local exclusive memory with shared
memory faster than any network. Or is this perhaps what they call a
supercomputer?

-- 
David Feuer
feuer@his.com
dfeuer@binx.mbhs.edu
Open Source: Think locally; act globally.


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/