2 Jul
2010
2 Jul
'10
1:35 a.m.
Eric Smith <eric at brouhaha.com> wrote:
Indeed. Which I think I tried pointing out.
Then you are saying that Don North's work on the CIS microcode for the
11/74 is a figment of his imagination? And the results they got back
Johnny Billquist wrote:
That might be correct.
Another way to name them would perhaps be:
KB11-B - Old 11/70 CPU with synch FPP.
KB11-C - New 11/70 CPU with asynch FPP.
KB11-CM - MP modified KB11-C
KB11-E(?) - The new 11/74 CPU with asynch FPP and CIS.
I seem to remember reading somewhere that the 11/74 CPU were to be
called KB11-E, but I also have this nagging feeling that KB11-E might
have been the 11/44, or possibly the 11/60.
The 11/44 CPU was a KD11-Z. The 11/60 CPU was a KD11-K. Now, as I
myself pointed out, RSX regards the 11/70mP as an 11/74, and
that is also what the CPU identification code in RSX calls it.
Since the 11/70mp and 11/74 were never official products, there was a
lot of conflation of the designations. Without the optional CIS,
software can't easily distinguish an 11/70mp from an 11/74, so it
probably simply didn't bother to try. Thus whether software reports the
CPU as an 11/70mp or an 11/74 doesn't really prove much of anything. But if we call
this an 11/74, what shall we call the 11/70 with CIS?
Fantasy? There wasn't such a thing, since there wasn't an 11/70 with a
KB11-E CPU that was necessary to accomodate the KE74-A CIS. from running performance tests on this hardware?
I think it's just easier, for this discussion, to call that the 11/74,
and call the multiprocessor PDP-11s that went out to field test, and
which also were kept running inside DEC until not long ago, 11/70mP.
Mind you - just for this discussion.
Otherwise I'm just happy to keep calling CASTOR:: an 11/74, just as my
emulated 11/74 (MIM::), which also don't have CIS...
Johnny