> As best as I can tell, you have to get a 400Mhz PII to match
> a 200Mhz PentPro.

Eeek.  This seems to go against a naive reading of Intel's published
figures.  Their UnixWare figures from 
http://developer.intel.com/procs/perf/PentiumII/spec95int.htm
seem to be similar to, but not as complete as their Windows NT
figures from
http://developer.intel.com/procs/perf/PentiumII/spec95int_win.htm
which state these results:

          Pro200       PII   233   266   300   333   350   400
SPECint95 8.09              9.38 10.70 11.90 13.00 13.90 15.80

Okay, so it's unfair cos the Pro200 there is a 256k cache model,
whereas the PIIs are all 512k, but from
http://developer.intel.com/procs/perf/archive/highend/spec95.htm
512k only brings the PPro up to 8.58.

So either:

	a) You're talking about a measure of performance that SPECint95
	   fails to measure

	b) You're talking about the 1Mb cache PPros that I can't find
	   on intel's web site (and are extremely expensive)

	c) Intel's compilers are supergood or cheating :-)

Which is it ?  If (a) does that then apply to mouse's application mix,
and if (b) then does that then applly to mouse's budget ?

Certainly to me the figures above seem to make the PII seem to me on
paper to be very attractive as integer compute per dollar boxes.  The
only reservation I have from what I see on paper is that is that the
commodity motherboards do not support memory interleave.

--