XPost: alt.folklore.computers
From: anton@mips.complang.tuwien.ac.at
Thomas Koenig writes:
>To be efficient, a RISC needs a full-width (presumably 32 bit)
>external data bus, plus a separate address bus, which should at
>least be 26 bits, better 32. A random ARM CPU I looked at at
>bitsavers had 84 pins, which sounds reasonable.
>
>Building an ARM-like instead of a 68000 would have been feasible,
>but the resulting systems would have been more expensive (the
>68000 had 64 pins).
One could have done a RISC-VAX microprocessor with 16-bit data bus and
24-bit address bus, like the 68000, or even an 8-bit data bus, and
without FPU and MMU and without PDP-11 decoder. The performance would
have been memory-bandwidth-limited and therefore simular to the 68000
and 68008, respectively (unless extra love was spent on the memory
interface, e.g., with row optimization), with a few memory accesses
saved by having more registers. This would still have made sense in a
world where the same architecture was available (with better
performance) on the supermini of the day, the RISC-VAX: Write your
code on the cheap micro RISC-VAX and this will give you the
performance advantages in a few years when proper 32-bit computing
arrives (or on more expensive systems today).
>So... a strategy could have been to establish the concept with
>minicomputers, to make money (the VAX sold big) and then move
>aggressively towards microprocessors, trying the disruptive move
>towards workstations within the same company (which would be HARD).
For workstations one would need the MMU and the FPU as extra chips.
Getting a company to avoid trying to milk the cash cow for longer
(short-term profits) by burying in-company progress (that other
companies then make, i.e., long-term loss) may be hard, but given that
some companies have survived, it's obviously possible.
HP seems to have avoided the problem at various stages: They had their
own HP3000 and HP9000/500 architectures, but found ways to drop that
for HPPA without losing too many customers, then they dropped HPPa for
IA-64, and IA-64 for AMD64, and they still survive. They also managed
to become one of the biggest PC makers, but found it necessary to
split the PC and big-machine businesses into two companies.
>As for the PC - a scaled-down, cheap, compatible, multi-cycle per
>instruction microprocessor could have worked for that market,
>but it is entirely unclear to me what this would / could
>have done to the PC market, if IBM could have been prevented
>from gaining such market dominance.
The IBM PC success was based on the open architecture, on being more
advanced than the Apple II and not too expensive, and the IBM name
certainly helped at the start. In the long run it was an Intel and
Microsoft success, not an IBM success. And Intel's 8086 success was
initially helped by being able to port 8080 programs (with 8080->8086
assemblers).
So how could one capture the PC market? The RISC-VAX would probably
have been too expensive for a PC, even with an 8-bit data bus and a
reduced instruction set, along the lines of RV32E. Or maybe that
would have been feasible, in which case one would provide
8080->reduced-RISC-VAX and 6502->reduced-RISC-VAX assemblers to make
porting easier. And then try to sell it to IBM Boca Raton.
An alternative would be to sell it as a faster and better upgrade path
for the 8088 later, as competition to the 80286. Have a RISC-VAX
(without MMU und FPU) with an additional 8086 decoder for running
legacy programs (should be possible in the 134,000 transistors that the
80286 has): Users could run their existing code, as well as
future-oriented (actually present-oriented) 32-bit code. The next
step would be adding the TLB for paging.
Concerning on how to do it from the business side: The microprocessor
business (at least, maybe more) should probably be spun off as an
independent company, such that customers would not need to worry about
being at a disadvantage compared to DEC in-house demands.
One can also imagine other ways: Instead of the reduced-RISC-VAX, Try
to get a PDP-11 variant with 8-bit data bus into the actual IBM PC
(instead of the 8088), or set up your own PC business based on such a
processor; and then the logical upgrade path would be to the successor
of the PDP-11, the RISC-VAX (with PDP-11 decoder).
What about the fears of the majority in the company working on big
computers? They would continue to make big computers, with initially
faster and later more CPUs than PCs. That's what we are seeing today.
- anton
--
'Anyone trying for "industrial quality" ISA should avoid undefined behavior.'
Mitch Alsup,
--- SoupGate-Win32 v1.05
* Origin: you cannot sedate... all the things you hate (1:229/2)
|
