Re: OT - cynical speculation

  The conspiracy theorem again?

  Well, this concept had considerable validity in respect of certain CPU
chips - as you would know - but it's only *slightly* likely in respect
of PICs.  The rationale for the '486 was that the co-processor was as
complex as the CPU proper, taking up half of the die area, and this
permitted "recovery" of 50% of the faulty chips - one with a faulty CPU
but functional co-processor was *not* salvageable.

  In regard to the PICs, I suspect the chip yield is so high that such a
scheme has no benefit.  And this is a very different end of the market -
making *cheap* chips rather than *premium* ones, such that aspects such
as packaging and inventory come into play.  For the wholesale price
difference between say, a 12C508 and a 12C509, and considering that the
volume market is quite likely in the lower-spec chip, it would hardly be
worth charging a premium for the higher-spec chip to have a split
production line (even the "tail end") and inventory.

  I think then, that you have to assume, the *only* reason to do this,
is if the lower-spec chip *really does* save - and substantially - on
silicon (die yield).

  I point out that this market (substantially automotive, might I
presume?) is *far* more competitive than the minicomputer market in the
'80s (wouldn't have been DEC, surely?).
-- 
  Cheers,
    Paul B.