On 08/06/2020 16:55, Chris Zach via cctalk wrote:
I have a MV1 board set here, I'd be willing to
fire up VMS to see if
it's slower if someone wants to send me a TK50 boot tape but I'm
guessing it would be the slowest Vax on earth.
Reasons why:
1) MV1 has only 4mb of memory to the 730's 5mb.
2) I believe the 730 accesses memory in 32 bit chunks as opposed to
the MV1 having to access memory over the Q bus in 16 bit words. So
you're double bucking there. Don't remember if the MV1 has a cache.
3) An 11/730 has that R80 controller. The R80 is most probably faster
than the RQDX3+RD54, note that the RQDX3 is a slug of a controller
compared even to a MTI ESDI controller. Possibly due to the MFM
interface.
4) The 11/730 could emulate pdp11 instructions, the MV1 could not.
Come to think of it I think the 730's floating point could do D,F,G,H
while the MV1 could only do D,F,G.
I'm not sure why they didn't do PMI memory on the MV1, possibly
because it's two boards already and they might already talk over CD
interconnect, limiting the memory to one card. But the result is a
really slow little VAX, probably slower than the 730.
I think the MicroVAX I was (at least in part) built quickly to prove
that it could be done and was always intended to be superseded by the
MicroVAX II (which took longer and was more efficient).
I'm sure I've seen explanations on the net before, but I can't find them
now. Maybe I read it on the EASYNET? In which case, until someone is
willing to release their backups of some of DEC's internal NOTES
conferences, we'll possibly never know.
Antonio
--
Antonio Carlini
antonio at
acarlini.com