So, where do
you draw the line?  I draw it at one piece of silicon that
 can stand alone with the usual support chips (ROM, RAM, glue chips). 
 I define a microprocessor as either a single chip processor, or as a
 small (to be defined ;-)) number of chips which are only ever used
 together to make a particular processor. 
 
Yes, I agree with that one.  Slight grey area here - things like the
floating point unit in the LSI11 or the CIS in the KDF-11 - are they
part of the microprocessor chipset or are they coprocessors?
  Byt that definition, anything built out of TTL, 2900
series, 3000 series,
 etc is _not_ a microprocessor. Those chips have uses other than for
 building that particular processor. But the F11 (PDP11/23) is - that
 chipset only ever gets used to make an 11/23 (or an 11/24 for the pedants
 :-)).
 So is the early IBM RISC 6000 processor. I seem to remember that's 8 or 9
 packages, but I would still class it as a micro.
 Other people insist a microprocessor is one chip only. I've got no 
So I notice.  But by that definition _none_ of the micro PDPs were true
microprocessors, or at least none up until the 73.  The Micro J-11
processor in the 73 was implemented as two chips on a large ceramic
carrier.  Was this also the case with later J-11s?
Philip.