🚀 go-pugleaf

In article <14873@agate.berkeley.edu>, Jonathan Dubman (128a-3aj@e260-3b.berkeley.edu) writes:
>So the Mac IIx has a 68030 and a 68882 running at the same speed as the
>original Mac II.
>
>BIG DEAL.

  I agree I'd like to see a BIG improvement in the Macintosh line, but
I can see what may be some of the reasons for not making a big jump
yet.

>I don't understand what Apple is doing here.  All the PC manufacturers are
>coming out with 20 and 25 MhZ processors with 30 MHz on the near horizon.
>And Apple comes out with a 16 MHz machine.

  Beware of comparing MHz.  Admittedly, this topic has been beaten to
death on various newsgroups, but processors vary a lot in what they
get done each clock cycle.  When you get up to 68020/68030-class
machines, the caching and prefetch start really making a difference in
terms of instructions-per-cycle.  Built in virtual memory is also a
real help, at least when you have a VM-based operating system--but it
slows down the processor a little bit.
  A 68020 (arguably) does more work each cycle than an 80286.  Also,
something else to watch is the memory access time.  With high-speed
memory prices going up, putting a 30-MHz 68030 in your machine will
cost a LOT--not just the processor and main memory, but putting a
reasonable cache (say, 32-64K) of really high-speed memory in to avoid
wait states (disclaimer: I haven't seen a 68030 hardware manual, so I
don't know its memory rate/cycle time ratio).

>So it has a built-in PMMU which
>future versions of the Mac OS will make use of, and UNIX can use, and it has
>all sorts of little optimizations, but really, why all the fuss over what
>seems like a very minor upgrade?

  I think the important thing here is the built-in PMMU.  Apple should
be getting a virtual memory system (not A/UX, but a *real* Mac one)
done fairly soon.  This will provide a box people can buy to run it
on, without needing an upgrade.  I suspect pretty soon the PMMU
upgrade will be standard equipment for a Mac ][.
  Also, the 68882 provides a lot of speed for some applications; Apple
might have decided not to wait on releasing the machine until they
could push up the clock speed some more if they thought the doubled
(or so) floating-point would broaden the market.

  Disclaimer: All of this is just my opinion.  I expect nobody else
              shares it.  I have no connection with Apple except as
              a customer.

+----------------------------------+------------------------+
| Anton Rang (grad student)	   | "VMS forever!"	    |
| Michigan State University	   | rang@cpswh.cps.msu.edu |
+----------------------------------+------------------------+