🚀 go-pugleaf

Joe Ragosta wrote:
> 
> In article <4d4omk$otj@steel.interlog.com>, daveacg@interlog.com (Dave
> Glue) wrote:
> 
> > On 11 Jan 1996 09:04:09 -0800, curt@beowulf.jpl.nasa.gov (Curt
> > Eggemeyer) wrote:
> >
> > >The Pentium Pro's maybe faster in some respect, but they don't compete
> > >with the powerPC chips when it comes to production costs. In the end
> > >Intel will have to reduce their massive transistor count if they hope
> > >to remain competitive with the other cpu producers.
> >
> > Really.  How do you know this?  Do you work on both lines?  Intel's
> > profit margins are VERY high.  For such high production costs, that
> > would seem strange, wouldn't it?
> 
> Well, PC Week just stated that Intel is releasing a chip code-named
> Klamath which will be a Pentium Pro w/o the cache module. This reduces the
> transister count by 75 or 80% if I remember correctly. So, Curt seems to
> be correct.
> 
> BTW, "sources" claim that Klamath will  still be faster than the fastest
> Pentium. No other performance details were available and Intel declined to
> comment on rumors.
> 

BTW, the 5.5M transistors in the PPro does not include the cache. The 
cache at 256K requires 15 million transistors, and 30 M for the 512K 
part.