4 Mar
2006
4 Mar
'06
5:10 p.m.
On my wandering around, I found the following claims on the Unisys
"history" web page (http://www.unisys.com/about__unisys/history/index.htm):
"1993 Unisys introduces 2200/500, the first mainframe based on
complementary metal oxide semiconductor (CMOS) technology."
Wasn't the ETA-10 CMOS-based? (Not to mention the whole subject of
"What's a mainframe?")
"1989 Unisys introduces Micro A, the first desktop, single-chip mainframe."
I don't know what to make of that claim. What's a mainframe. From what I
can gather, the Micro A was pretty much a plug-in coprocessor PCB in an
otherwise x86 box that booted OS/2.
Comments? Anyone on the list a Blue Bell veteran who can give the real
scoop on the Micro A?
Cheers,
Chuck
4 Mar
4 Mar
5:55 p.m.
"Chuck Guzis" <cclist at sydex.com> wrote:
On my wandering around, I found the following claims
on the Unisys
"history" web page (http://www.unisys.com/about__unisys/history/index.htm):
"1993 Unisys introduces 2200/500, the first mainframe based on
complementary metal oxide semiconductor (CMOS) technology."
Wasn't the ETA-10 CMOS-based? (Not to mention the whole subject of
"What's a mainframe?")
But the ETA-10 was by very few standards a "mainframe". It was clearly
a supercomputer. Not aimed at (for example) payroll by any means!
"1989 Unisys introduces Micro A, the first
desktop, single-chip mainframe."
I don't know what to make of that claim. What's a mainframe. From what I
can gather, the Micro A was pretty much a plug-in coprocessor PCB in an
otherwise x86 box that booted OS/2.
As to "what's a mainframe", as Unisys uses the term it's a computer
used for "traditional" business purposes (financial/payroll/purchasing/
invoicing), probably running dusty-deck COBOL programs.
Tim.
6:14 p.m.
On 3/4/2006 at 5:55 PM shoppa_classiccmp at trailing-edge.com wrote:
As to "what's a mainframe", as Unisys
uses the term it's a computer
used for "traditional" business purposes (financial/payroll/purchasing/
invoicing), probably running dusty-deck COBOL programs.
That could mean any minicomputer and quite a few micros of the same era.
Certainly COBOL was available for the x80 by then and many micros were
running database and the usual finiancial apps, including payroll,
inventory, AR, AP and GL.
But saying a supercomuter isn't a mainframe is bending things quite a bit,
don't you think?
Cheers,
Chuck
7:04 p.m.
William Donzelli <aw288 at osfn.org> wrote:
Gotta remind me what ALL the tests are :-). I agree with
you that a super is not a mainframe, but cannot tell you
all the reasons why!
S = Storage maybe? Or maybe not,
certainly many of the CDC and CDC-derived supers did have channel-type
processors for peripherals, and in that way they were like mainframes.
And on a certainly tangential subject: any good books/articles
giving a historical retrospective to the "Fifth Generation" computers
of the 80's?
Tim.
But saying a
supercomuter isn't a mainframe is bending things quite a bit,
don't you think?
They tend to fail on the R, A, and S portions of the test. 
6:10 p.m.
From what I
can gather, the Micro A was pretty much a plug-in coprocessor PCB in an
otherwise x86 box that booted OS/2.
"1989 Unisys introduces Micro A, the first
desktop, single-chip
mainframe."
Lucky Me, somebody turned up some history on my oddity
There is in fact a desktop emulator of a Unisys mainframe
that boots the coprocessor from within OS/2. I have one
here. Happened across it by pure luck. I'm not an expert
in the MCP O/S so have been in lurkmode on this for a
longnow while.
John A.
end of line
5 Mar
5 Mar
1:57 p.m.
On 3/4/2006 at 6:10 PM John Allain wrote:
Lucky Me, somebody turned up some history on my oddity
There is in fact a desktop emulator of a Unisys mainframe
that boots the coprocessor from within OS/2. I have one
here. Happened across it by pure luck. I'm not an expert
in the MCP O/S so have been in lurkmode on this for a
longnow while.
From what I've been able to tell, the thing
executes a superset of the old
1107 instruction set. Given the date of introduction
(1989), it's got to
be one of the last holdouts for non-power-of-2 word length (36 bits). An
odd bird indeed--OS/2 sounds as if it was used to handle I/O for the Micro
A.
Cheers,
Chuck
7:25 a.m.
Chuck Guzis wrote:
Wasn't the ETA-10 CMOS-based? (Not to mention
the whole subject of
"What's a mainframe?")
Out of interest - how did the term come into being? When was it first used and
why? I don't think I've ever seen the reason given, and the actual term seems
a little strange when thought about...
cheers
Jules
8:29 a.m.
Jules Richardson <julesrichardsonuk at yahoo.co.uk> wrote:
Chuck Guzis wrote:
More generically in the electronics/telco industry, a "mainframe" is
the unit into which modules are plugged. In the case of the tube
computers of the 50's, the modules might have half a dozen or so tubes,
weigh 5 or 10 pounds, and be a foot long and a 4 inches high and a few
inches wide. And it would take hundreds/thousands of the modules
(each being a couple of gates or a flip-flop) to implement a computer.
Different technologies have different concepts of "modules" (some will
be smaller, others will be larger) but the thing they all
plug into, if big enough, is the "mainframe".
Sort of the same concept as a "backplane" in a 70's era mini, but
the "mainframe" was spread out over many racks, used discrete wires, etc.
Tim.
Wasn't the ETA-10 CMOS-based? (Not to
mention the whole subject of
"What's a mainframe?")
Out of interest - how did the term come into being? When was it first used and
why? I don't think I've ever seen the reason given, and the actual term seems
a little strange when thought about...
6714
days inactive
6715
days old
10 comments
5 participants
participants (5)
-
allain＠panix.com -
aw288＠osfn.org -
cclist＠sydex.com -
julesrichardsonuk＠yahoo.co.uk -
shoppa_classiccmp＠trailing-edge.com