What's the rarest or most unusual computer-related item

Jay West jwest at classiccmp.org
Mon Jan 16 19:23:21 CST 2017

Mark wrote...
Was it possible to configure an Access system with a mix of a 21MX and 2100?

(I'm not challenging the assertion; it just never occurred to me...)
According to the documentation - specifically "no". Both processors must be
the same type.
However, after digging in to it year ago, I see no reason that it shouldn't
work and others on the list said they were fairly certain that it did work.

Actually, the thing I'm calling a 21MX is listed in the components list as a
"1000".  Perhaps this implies that it (or they) was running one of the RTE
operating systems.
Most of the designations were pure marketingspeak, and on top of that who
knows what happened in shops out in the field. Operators often did things
that were perfectly ok but not sold that way.

In the end.... HP1000 meant a 21MX M, E, K, or F - which was 2105, 2108,
2112, 2109, 2113, 2111, 2117. I think some later AXXX systems were also
considered HP1000's, but that's out of my area (too new). 

Hard to tell just by looking if that machine is an HP2000 system. Some
indicators to look for:
Dual Processors
12920/21 (or the older mux set) boards and the processor interconnect cables
By studying what boards are in which slots in both processors
IOP firmware on a FAB (can't be in a FEM I don't believe, due to hard coded
I/O addresses)
Must have mag tape, cannot run without it.
Any HP-IB devices is a certain clue that it's NOT an access system (nor any
Hp2000 system).

But the above is just best guess (other than the IOP microcode)... because
there are plenty of examples of those items in non-Access HP2000 systems and
even NON-hp2000 systems. 2000E used only one cpu (and I think another prior
system as well) for example. Dual cpus was written up by HP as "heres how
you can slave an additional cpu for special applications", etc.

So the best way is to take a snapshot of both rear card cages showing the
extractor handles and I can then most likely determine with a fair (but not
conclusive) amount of certainty.


More information about the cctalk mailing list