• Please review our updated Terms and Rules here

*** Happy 50th, System 360! ***

1944GPW

1944GPW

Veteran Member
Joined
Apr 26, 2011
Messages
813
Location
Brisbane, Australia
Today 7th April marks the 50th anniversary of the announcement of the IBM System 360 range, an architecture and hardware platform concept that revolutionised the industry.
http://www.usatoday.com/story/news/nation/2014/04/06/ibm-mainframes-mark-50th-anniversary/7364535/

Personal connection- My dad was an Australian IBM CE who was sent on assignment with family to Hursley UK in 1964 then Poughkeepsie in 1965 to learn the hardware of the Model 40 (2040 CPU) so that he'd be able to work on the /40's to be installed around the Asia/Pacific region in the 60s, and also train other CE's. I'm not sure how many people can claim that a computer influenced where they were born, but that's the case for me - because of the 360/40 I was born near IBM's Hursley labs whilst my parents were there. (I'm heading off camping for a few days right now but will post some more when I get back.)

Steve.
 
Oh shoot... we should have all brought some of our 360's to work for a party. ;-)
 
Yes, but please no 360/20 systems--I consider those to be the idiot offspring.

Code: 
        BALR    15,0
        USING   *,15

The longevity of the S/360 architecture should be ample rebuttal to those who think every computer system should implement a hardware stack.
 
I just had to read the first headline I saw about this birthday, "The mainframe turns 50" since the first mainframe I saw in person was in 1958, and it had been there for a while :)
 
There's also a Register article that written a bit better here. But it still has a few things wrong with it.

One thing that I think that almost everyone misses is that IBM had a strategy to use the S/360 platform to replace most of its second-generation workhorses. A substantial part of the installed base of the S/360s spent a fair amount of time in emulation mode. I can think of a military installation where 4 S/360 65s ran in 7080 emulation mode the entire time. (The customer had tons of COBOL code that had, over the years been binary-patched repeatedly with 7080 Autocoder--and never documented). 1401 emulation was hugely popular, as was 7090. I believe I recall a report of one installation that ran 650 emulation inside of 1401 emulation on a S/360 box.

(Again, I don't count the lowly model 20) The programming support was vast, from lowly BPS, requiring 8K and a tape drive. to the high-end timesharing versions. Yes, IBM spent a lot of money on the projects, but reaped vast profits.

The architecture was novel enough to be imitated by several vendors. Besides, who could resist a system where one assembly mnemonic was "ZAP"?
 
Doug G said:
I just had to read the first headline I saw about this birthday, "The mainframe turns 50" since the first mainframe I saw in person was in 1958, and it had been there for a while :)
Click to expand...
The System/360 is considered to be the first "modern" mainframe computer for several reasons but the major one is that it was the first computer to divorce architecture from implementation.

It also introduced the 8-bit byte standard - and where would we be without that convention!

It's interesting to hear of someone connected to the Model 40. It was the first /360 to be powered up in the world, at IBM Hursley Park.

There are some further links here: http://hursley.slx-online.biz/
 
Last edited:
Beerhunter said:
It also introduced the 8-bit byte standard - and where would we be without that convention!
Click to expand...

Probably using 12-bit characters, I imagine (yes, there were systems that did that).

What I find more interesting is that the S/360 ushered in the convention of having word lengths be a power of 2. (8, 16, 32, 64...) Before that became accepted practice, the more common convention was to use a multiple of 6 bits.

Having character addressability was inherited from the IBM business world and preceded the S/360 In scientific computer architectures, simply using word addresses was considered to be sufficient--and in many cases was perfectly adequate. But it seems that after S/360, character addressability was considered to be a "must" in an architecture.
 
Myself (left) and my brother at the console table of a Model 40 in April 1967, at IBM in Kent Street, Sydney, Australia (in what was the landmark 'pagoda building' just on the CBD side of the harbour bridge). This would have been on a weekend and my dad would have been in there to work on that machine at a quiet time no doubt :) Notice on the panel with the four knobs immediately behind my head (Console panel 'C') there is a bank of switches all in the up position except for a black one down. This is the DISBL INTVL TIMER switch which in that position turns off the system timer so that doesn't interfere with the currently running program. I guess my dad had been running some diagnostics or some such, I don't know.
Myself_and_brother_System360Mod40console_IBM_Kent_Street_Sydney_April_1967_2.jpg

Myself_and_brother_System360Mod40console_IBM_Kent_Street_Sydney_April_1967_1.jpg
Somewhere my mum has some slides my dad took at IBM Hursley but I still haven't found them yet. Thanks for the link Beerhunter, I would surely visit your museum if I'm in the UK someday!

Steve.
 
Last edited:
You must log in or register to reply here.
Back
Top