🚀 go-pugleaf

In <3vkii2$o6o@freud.inst.com> gato@freud.inst.com (gato) writes:
>: Infoworld-- PowerMac is almost 2x Pentium speed.
>: PCWeek (no Mac friend, BTW) -- PowerMac is 2x Pentium speed.
>: MacWorld -- PowerMac is 2x Pentium speed.

> [munch]

>Finally, the P6, whether or not the initial 133MHz version's stats impress
>you, will be in an architecture designed around SMP.  Systems with as many
>as four procs are expected.  Even Windows 95 would run quite nice with
>that kind of power, if the fools at Microsoft had bothered to add
>multithreading to the PEM.  I guess you're looking at about 800 SPECint92
>there, and that's just ihe 133MHz chip.  The 200MHz chips, of course, will
>blow that out of the water.  In other words, Windows NT is still around
>for a reason: it will become the standard OS for these SMP systems (though
>Solaris would be a better alternative;>)

I see no architectural reason why the PPC can't be used to build SMP boxes
in exactly the same way that there are multiprocessor x86 machines,
multiprocessor Alpha machines, multiprocessor R4k machines..

Frankly, I'm not that impressed with the P6's figures. It's offering about
the same SPECint/Mhz that the PPC is offering NOW (a la Powermac 9500).
And Intel don't seem to be saying what the SPECfp is going to be, which
doesn't fill me with confidence!

Not that I want to use a Mac... An SMP NT box will do me just fine :)

/\/.


-- 



|\Nigel Magnay\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\[_][~][X]|
|magnayn@cs.man.ac.uk       {Main}|   3rd yr. CS, University of Manchester   |
|nigel@rcp.co.uk      {Summer job}|                                          |
|http://www.cs.man.ac.uk/~magnayn |"Because you know I'm better than you are"|
-----------------------------------------------------------------------------/