[Simh] HP 2100/1000 A-series

[Holger Veit]

Am 07.06.2015 um 05:40 schrieb J. David Bryan:
> On Saturday, June 6, 2015 at 21:44, Cory Smelosky wrote:
>
>> It seems there's no current SIMH support for the HP 1000 A-series system
>> (A900, A990, A600 et cetera) meaning RTE-A (with the ARPA/1000 software!)
>> is unfortunately out.
> That's correct.
>
>
>> Looks like enough docs for at least some of the later ones are up on
>> Bitsavers...any plans for implementing these missing models?
> Holger Veit <holger.veit at iais.fraunhofer.de> had been looking into doing a
> SIMH A-Series simulator several years ago, but I've not heard what progress
> he made, if any.
>
> Regrettably, relatively little of the existing HP 1000 M/E/F-Series
> simulator could be leveraged for an A-Series simulator.  The base set
> instructions are the same, although the I/O instructions and I/O structure
> are entirely different, and there are a number of new instructions.  Also,
> essentially none of the M/E/F peripherals (disc, tape, terminal
> multiplexers, line printers) were supported on the A-Series, so a whole new
> set of peripheral simulators would have to be written.  An off-the-cuff
> guess would be that reusing the base set would cover maybe 20% of the full
> simulator code.  The other 80% would have to be written from scratch.
>
> On the plus side, the full RTE-A sources, the A-Series microcode, and
> several RTE-A binary releases are available in the HP 1000 Software
> Collection on Bitsavers, so the job is by no means impossible.
JDB summarizes the problem quite precisely.

After the microcode ROM stuff, JDB and I added to the HP1000 system, I 
infact looked at the A series,
to possibly extend the simulator with "the missing A instructions". It 
turned out that the A series is
not only a class of quite different systems with entirely different 
properties but also with a completely new
I/O system (based on a monolithic chip). It is a new platform. Workload 
prevented me from following this
further, but I am still attracted by such a simulator, mainly because of 
the microcode which is well documented
for some of the models, as well as for RTE-A itself which is a 
technological progress to the older RTE-IV/6VM systems.

On the documentation side, the I/O chip is well documented, however the 
various peripherals that are built around it
are less described. That time I didn't find a complete enough set to 
combine a useful A hardware. The CPU part itself
is likely the smallest problem even if I'd start this from scratch as 
well (as a plain microcode interpreter) rather than
lifing the base instructions from the HP1000 emulator.

The things are not forgotten, as well as the PDQ3 and Sage-II/IV things 
- it's just the general problem that the day is
limited to 24h.

-- 
Holger

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 6161 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://mailman.trailing-edge.com/pipermail/simh/attachments/20150608/384142b9/attachment.bin>