16 Feb
2015
16 Feb
'15
8:12 a.m.
On 2015-02-16 09:42, Eric Smith wrote:
On Mon, Feb 16, 2015 at 1:23 AM, Paul Birkel
<pbirkel at gmail.com> wrote:
Agree.
Ah. I need to count better. There are *two*
MKA11 per CPU, thus each only
*needs* just a 1 Mw configuration due to address-space limitations.
I don't think so. The system shown in the photos has four CPUs, and
four MKA11 boxes. The MKA11 boxes are located in the right half of
the cabinet housing CPU C, and the left half of the cabinet housing
CPU D. There's no evidence in the photos of there being any more
MKA11 boxes other than those. For instance, the right half of the
cabinet housing CPU B clearly contains no MKA11 boxes, but rather what
appear to be Unibus expansion boxes. Only one photo shows the the
right half of the cabinet housing CPU A open, and the view is
partially obstructed by the CPU being extended out on its rails, but
it definitely does not appear to have any MKA11 boxes either. If each MKA11 box has eight MS11-KE storage array
modules (M7984),
that is 256KW per MKA11, for 1MW total. While that's only just over
half of the physical memory address space of the CPUs (1920KW, because
the top 128KW is reserved for the Unibus address space), that seems
like a very plausible real-world system configuration.
Box 0 and 2 are clearly setup to be interleaved with each other, and
both have a 0 starting address, is is located in the cabinet holding CPC.
I wish I could read out what the thumbwheel switches on box 1 and 3 are
set to in the cabinet holding CPD, as that would tell us how much memory
are in each box.
But I would have expected the boxes to hold close 512KW per box. With
the 128KW modules, that would be easy to reach.
Johnny