13 Dec
2001
13 Dec
'01
12:19 a.m.
I remember reading that a principal difference between a microcomputer, back in
the old days, and a minicomputer, was that the mini had an asynchronous bus.
That was, ISTR, a definition based largely on the DEC-realized version of the
minicomputer.
If it topped out at 7 MBps, it was probably because the bus handshake was
clocked with a CPU clock, in order to ensure the CPU would "see" the
transitions.
Dick
----- Original Message -----
From: "Dave McGuire" <mcguire(a)neurotica.com>
To: <classiccmp(a)classiccmp.org>
Sent: Wednesday, December 12, 2001 10:55 PM
Subject: Re: 1GHz PDP-11 (Re: Prints for an 11/70)
On December 12, Jeffrey S. Sharp wrote:
What little I've read about UNIBUS told me
that UNIBUS has no set clock
rate, and that the speed of communication between two devices would be the
the highest rate that both devices could handle. If you've got nothing
but Sridhar-made fast devices on the bus, what stops you from having a
UNIBUS operating at say, 33MHz to 100MHz on average?
I don't recall the specifics of Unibus...but its bandwidth is commonly
stated as being about 7MB/sec. If that's the case, then it's unlikely
that it's asynchronous. But perhaps it is asynchronous, and 7MB/sec
was just the maximum.
-Dave
--
Dave McGuire
St. Petersburg, FL