etc.

gjphw@ihuxm.UUCP (03/21/84)

   Most people probably know this, but I thought it might be nice to get this
 into the open.  The greatest value for a 64 bit microprocessor chip, as the
 Japanese have announced an intention to develop, is to the science and
 engineering communities.  IBM and their 32 bit architecture dominate the
 market because of the needs of business, and the fact that the business end
 makes the purchasing decision.  An IBM mainframe is not particularly well
 suited for scientific computing, and a 64 bit micro, included in a processor
 array with floating point support, would be ideal for serious numerical work.

   Think of it as a CRAY in a breadbox....
-- 

                                    Patrick Wyant
                                    AT&T Bell Laboratories (Naperville, IL)
                                    *!ihuxm!gjphw

bcw@duke.UUCP (Bruce C. Wright) (03/23/84)

That's just it - there might be some relatively small market for a
64-bit micro, but it's doubtful that it will ever be commercially
viable in the sense that the IBM-PC or the Apple were/are.  The
vast majority of users could hardly care less about complex numerical
work - they just want to edit letters, run spreadsheets, etc., none
of which particularly need either 64-bit registers or a 64-bit
address space.  I suspect that most of the 64-bit designs which
may come out will be like the Cray:  expensive relative to other
machines, and devoted primarily to special-purpose uses.  Nobody
denies that there *are* such special-purpose uses, but they are
probably less than 0.1% of the entire computing scene.


			Bruce C. Wright

els@pur-phy.UUCP (Eric Strobel) (03/23/84)

      A small market???  Do you realize how many physicists, chemists,
engineers (of all types), architects, etc. there are??  Literally into
the millions!  This is a number roughly comparable (or perhaps larger)
than what I would guess the number of businesses to be.  I can't see
a company turning down a potential market like that!!  Besides, I'm sick
of micros being taylored to businesses!  I want something USEFUL (my standards
of course)!





       (`')           (`')
         \\   _____   //           Writing cause I got work, hanging by
          \\ /     \ //             my bruised ,bleeding and mangled thumbs
           \/ O   O \/              at the off-the-wall teddy bear keyboard of
            |   o   |
             \_____/                ERIC STROBEL
             /|+++|\
            //-----\\               decvax!pur-ee!Physics:els
           //       \\
         (_^_)     (_^_)

azia@utzoo.UUCP (Anton J Aylward A/S) (05/03/84)

Back in the days when "micros" was a term used by fashion writers,
and the 8080/6800 were so new you had to have an inside line
to get any solid info on how to make them work, I was playing round
with 'bit-slice'.

Bit-slice was new then, as well.  National and AMD had their slice
chips, Intel had a something they couldn't have like 'cos it didn't
survive as the 2900 and IMP families did.

For my sins, (looking back, I think this was one of them) I had the
misfortune to wirewrap a set of those slices, together with LOTS of
random logic ('cos there weren't the slice support chips) to emulate
a NOVA.  (The computer, not the car !) Why a NOVA ?  Cos some idiot
before me had worked out the microcode.