Wayne Chirnside wrote in a message to Roy J. Tellason:

WC> Saw a Gigabyte board on the net with every imaginable port and it
WC> had either 5 or six available slots, supported _4_ Gig of RAM and I
WC> believe a 4 Gig intel processor!!!

RJT> Only one?  :-)

WC>  Yeah  ain't it a bitch 

Yeah.  :-)



RJT> Cool.  I remember looking at one MB a few years back which could 
RJT> hold two CPUs,

WC> I wouldn't know how to apply them and nothing I'm currently doing 
WC> would require 2 processors.

Even the first version of Slackware I installed had SMP support,  if you
selected the right kernel.    You'll hit a point sooner or later
where things get *really* bogged down...

RJT>  and in addition to the usual stuff had _three_ SCSI channels on board,
RJT> including one wide one.  Only problem I had with it was that it wanted
RJT> P-II (or Celeron) type CPUs,  and I didn't care to pursue those,

WC> Neither do I but that's what the H.P. has and it seems fine to
WC> me. I AM looking to see what I might replace it with.

Unfortunately it seems that once you get past Socket 7 the two major mfr's
diverge quite widely,  and there's little interchangeability between them.

RJT> preferring AMD parts for the most part.  This was available for $75 or
RJT> so,  back about 3 years ago.

WC> Things have changed enormously.

No surprise there.

WC> Intel has radically cut their processor costs except at the 
WC> bleeding edge, say like 2.8 Ghz on up. 

They up that high already?  Jeez.

WC> Last I looked you could get an Intel 2.5 GHz for something like 
WC> you're quoting for the AMD or a bit more.  I suspect Intel cut it's 
WC> margin as AMD was getting innovative and becoming a real threat in 
WC> hopes of forcing them out of the marketplace.

It ain't gonna happen.  They've been playing this game for a long time...

WC> Intel has the name recognition 

Yeah,  "Intel inside" is a warning label!  :-)

WC> and if they keep this price strategy up long enough they have the 
WC> resources to outlast AMD and corner the market.

It ain't gonna happen.

WC> I prefer healthy competition and the best for AMD as it makes for 
WC> cheap CPU's for us be they AMD or Intel!

Yep.

RJT> I can still remember a few years back how this guy I knew had a
RJT> 386 MB that had a bad problem in the IDE part.  He could disable
RJT> it in CMOS,  but it still wouldn't let him plug in another "paddle
RJT> card" to use instead.  That MB was of _very_ limited usefulness
RJT> after that.  If you can't take it out...

WC> I suspect an EideMax or Pro card would have worked fine there as
WC> in stock configuration it takes over from the existing controller.

RJT> But the existing controller was bad,  and wouldn't let anything else
RJT> work in there.  I dunno...  That incident sorta put me off on having
RJT> lots and lots of stuff built in.

WC> If the board was old disable via jumper and use a card, if the
WC> board was newer like my H.P. there's an option to completely 
WC> disable in CMOS. YMMV as I've never lost a controller card of any 
WC> type.

I never had something built-in like that go bad on me either,  but this guy
was into a *lot* of messing around with hardware,  always swapping stuff
around,  adding and removing bits,  etc.

---