🚀 go-pugleaf

Andy wrote:
> 
> On Fri, 06 Jun 1997 11:22:03 -0700, Levi Cochran
> <levi.c...@Ebay.sun.com> wrote:
> 
> > I read about this computer in the paper.
> > Isn't the first computer to do a trillion calculations a sec.
> > I think it had 200+ PPros in parallel.
> >
> > All of this could have actually been a dream I had, but I think it's
> > pretty close to being correct.
> 
> 200+ PPros in parallel would not make the fastest computer.  Cray
> makes supercomputers using 512 500MHz Alphas in parallel.  To make an
> Intel-based machine as powerful as this, you'd need to parallel 2048
> PPros (large-scale parallelling is done in powers of 2).  This begs
> the question: why?  The Alpha solution would be more efficient and
> less expensive to implement.
> 
> I could see a pretty powerful computer being put together with PPros,
> but though I can't vouch for whether or not it has been done on this
> scale, I could think of better ways of getting that kind of power.
> 

A complete hardware breakdown of the system was given in the magazine
BYTE last year in an article about the world's fastest 5 machines. It
was astonishing and there were thousands of PPros in the Intel machine.
For example each IO controller had it's own bank of 4 PPros, and there
were dozens of these allowing multi terabyte fast disk access.

There may be better ways of getting more power, maybe cost came into the
equation since most components were available of the shelf, but as it
stands the machine is the world's fastest by some margin.

If anyones interested I can get more details.

--Pete--