6 Nov
2017
6 Nov
'17
8:52 p.m.
Hi all --
I mentioned a few weeks back that I picked up an HP 9836CU workstation.?
The "U" variant differs from the normal 9836 in that it contains an
upgraded CPU board that allows it to run early versions of HP-UX.? (The
"C" indicates that this machine has a color display, which is also cool
but not what I want to talk about here today.)
Information on the Internet varies about what microprocessor the 9836U
actually contains -- some sources say it's a 12Mhz 68010, some say it's
a 12Mhz 68000.? I found some internal HP marketing text that
corroborates the straight-68000 story.? I'd link it here, but
hpmuseum.net appears to be having technical issues at the moment.
My 9836CU has the 12Mhz 68000 (HP internal part number 1820-3288) fitted
in a socket on an 09836-66511 board (with the expected 16K SRAM cache
and MMU logic) and the processor is identified at power-up as a 68000.
Just for fun, I swapped the processor with a PGA 68010 and it powers up
and runs just fine, and identifies the processor as a 68010.? (Still
won't boot the copy of HP-UX 5.0 on the hpmuseum site, though...)
I'm curious if other people out there with 9836U's can confirm whether
their machine has a 68000 or a 68010 in it, I'd just like to settle the
internet discrepancy once and for all :).
- Josh
6 Nov
6 Nov
8:56 p.m.
The hpmuseum.net is indeed having technical issues at the moment. The after effects of an
Opsys upgrade to the web server it seems.
Service will be restored as soon as possible!
David Collins
On 7 Nov 2017, at 3:52 pm, Josh Dersch via cctalk
<cctalk at classiccmp.org> wrote:
Hi all --
I mentioned a few weeks back that I picked up an HP 9836CU workstation. The
"U" variant differs from the normal 9836 in that it contains an upgraded CPU
board that allows it to run early versions of HP-UX. (The "C" indicates that
this machine has a color display, which is also cool but not what I want to talk about
here today.)
Information on the Internet varies about what microprocessor the 9836U actually contains
-- some sources say it's a 12Mhz 68010, some say it's a 12Mhz 68000. I found some
internal HP marketing text that corroborates the straight-68000 story. I'd link it
here, but hpmuseum.net appears to be having technical issues at the moment.
My 9836CU has the 12Mhz 68000 (HP internal part number 1820-3288) fitted in a socket on
an 09836-66511 board (with the expected 16K SRAM cache and MMU logic) and the processor is
identified at power-up as a 68000.
Just for fun, I swapped the processor with a PGA 68010 and it powers up and runs just
fine, and identifies the processor as a 68010. (Still won't boot the copy of HP-UX
5.0 on the hpmuseum site, though...)
I'm curious if other people out there with 9836U's can confirm whether their
machine has a 68000 or a 68010 in it, I'd just like to settle the internet discrepancy
once and for all :).
- Josh
8:59 p.m.
On Tue, Nov 7, 2017 at 4:52 AM, Josh Dersch via cctalk
<cctalk at classiccmp.org> wrote:
I'm curious if other people out there with
9836U's can confirm whether their
machine has a 68000 or a 68010 in it, I'd just like to settle the internet
discrepancy once and for all :).
Mine identifies the CPU as a 68010 in the power-on diagnostic. But from what
I remember the PGA socket could also take a 68012 (with extra address pins
brought out). I don't have such a chip, so no idea what it would identify as.
-tony
10:39 p.m.
On Mon, Nov 6, 2017 at 9:59 PM, Tony Duell via cctalk <cctalk at classiccmp.org
wrote:
Mine identifies the CPU as a 68010 in the power-on
diagnostic. But from
what
I remember the PGA socket could also take a 68012 (with extra address pins
brought out). I don't have such a chip, so no idea what it would identify
as.
Do you mean that it will actually use the extra address pins?
I suppose the most likely way for the software to identify the MC68012 (vs
MC68010) would be to try accessing two memory addresses differing only in
address bits A24 or higher (but not A30), and test whether whatever MMU
hardware they've built will actually map them distinctly. The other
difference in the MC68012 is the availability of a /RMC pin to better
identify read-modify-write cycles, but since their board has to work with
an MC68010, I doubt that it would use the /RMC signal at all.
I've heard claims that HP used the MC68012 in some systems, but I've never
seen any definite confirmation.
7 Nov
7 Nov
1:34 a.m.
On Tue, Nov 7, 2017 at 6:39 AM, Eric Smith <spacewar at gmail.com> wrote:
On Mon, Nov 6, 2017 at 9:59 PM, Tony Duell via cctalk
<cctalk at classiccmp.org> wrote:
Well, theyare certainly wired to the MMU circuitry. I have no idea
if any software makes use of it though.
-tony
Mine identifies the CPU as a 68010 in the power-on diagnostic. But from
what
I remember the PGA socket could also take a 68012 (with extra address pins
brought out). I don't have such a chip, so no idea what it would identify
as.
Do you mean that it will actually use the extra address pins?
11:05 p.m.
On 11/6/2017 8:59 PM, Tony Duell wrote:
On Tue, Nov 7, 2017 at 4:52 AM, Josh Dersch via
cctalk
<cctalk at classiccmp.org> wrote:
Interesting.? So it looks like either processor was possible for the U
variant.? (With a 68012 as a possibility as well.)
hpmuseum.net is back up; here's the document I was referring to that
mentions the 68000, see page 16:
http://www.hpmuseum.net/pdf/ComputerNews_1983_Dec1_37pages_OCR.pdf
What BOOTROM version do you have in your machine?? I have 4.0, I'm
curious if there's a later version.
Thanks,
Josh
I'm curious if other people out there with
9836U's can confirm whether their
machine has a 68000 or a 68010 in it, I'd just like to settle the internet
discrepancy once and for all :).
Mine identifies the CPU as a 68010 in the
power-on diagnostic. But from what
I remember the PGA socket could also take a 68012 (with extra address pins
brought out). I don't have such a chip, so no idea what it would identify as.
-tony
11 Nov
11 Nov
9:05 p.m.
On 11/8/2017 2:31 AM, Tony Duell wrote:
On Wed, Nov 8, 2017 at 7:05 AM, Josh Dersch
<derschjo at gmail.com> wrote:
Found a doc on Bitsavers
(http://bitsavers.org/pdf/hp/9000_200/9000-200_periphSupp_Dec89.pdf, see
page 4-2's footnotes) that indicates that there were two variants of the
9836U -- the earlier with the 12.5Mhz 68000, the later with a 68010.?
Apparently the processor difference wasn't enough to require a model
designation of its own.
Apparently what I want for my system, HP-UX-wise is version 2.1. 5.1
just hangs after loading the kernel (no diagnostics) with either
processor, I assume the processor board itself must have some revisional
differences.? Maybe someday I'll track that down...
- Josh
On 11/6/2017 8:59 PM, Tony Duell wrote:
What BOOTROM version do you have in your machine? I have 4.0, I'm curious
if there's a later version.
Mine is 4.0
-tony
16 Nov
16 Nov
8:11 p.m.
On Mon, Nov 6, 2017 at 8:59 PM, Tony Duell <ard.p850ug1 at gmail.com> wrote:
On Tue, Nov 7, 2017 at 4:52 AM, Josh Dersch via
cctalk
<cctalk at classiccmp.org> wrote:
I picked up a 68012 (cheap) and stuck in in the 9836U this evening. It
works, and is identified as a 68012 during power-up diagnostics.
So now we know. Now what am I gonna do with all that address space? :)
- Josh
I'm curious if other people out there with
9836U's can confirm whether
their
machine has a 68000 or a 68010 in it, I'd
just like to settle the
internet
discrepancy once and for all :).
Mine identifies the CPU as a 68010 in the power-on diagnostic. But from
what
I remember the PGA socket could also take a 68012 (with extra address pins
brought out). I don't have such a chip, so no idea what it would identify
as.
-tony
2592
days inactive
2602
days old
8 comments
4 participants
participants (4)
-
ard.p850ug1＠gmail.com -
davidkcollins2＠gmail.com -
derschjo＠gmail.com -
spacewar＠gmail.com