back to article Sputnik, spaghetti and the IBM SPACE machine

The fall of 1957 was a low point not only for the United States, but for another high-minded world power: IBM. As the US looked up at Sputnik, without a satellite of its own, IBM was haunted by a rival machine known as the Gamma 3. Built by an upstart French outfit, the Gamma had trumped Blue Blue's fledgling computer tech - …

COMMENTS

This topic is closed for new posts.
Anonymous Coward

The most advanced feature...

...of the 1401 was the 1403 chain printer. It was fast, quiet and reliable and produced good clear output. It was still being used in the 1970s on the 360 and 370 series. With it companies could reliably produce cheques and invoices of a consistent quality which could be sent to a customer.

The competing manufacturer I worked for at the time couldn't produce a printer that didn't produce wavy lines until the mid 1970s. Our printers were very fast but didn't produce the quality to send out as commercial documents.

The 1401 processor itself wasn't particularly advanced and was not so different from what was produced by other manufacturers. That came later with the 360 series. But put it together with the excellent card, tape and paper handling capabilities and the 1401 became a winning system.

0
0
Thumb Up

Ah..Ha! He worked for ICL!!

I had some experience of 1401 computers but from afar. I agree about the print out. Like the tractr1 printers on the System /3 they were responsible for many multipart documents, clean and easy to read.

I had a brief stint on ICL in the 80's - How does a firm who came up with the DDE and bought Singer computers not be able to produce a data entry system which submitted jobs?? Or a decent printer?

I remember all the bootlace like wires for the 1401 and the stories of folks who wrote sort routines...

Those were the days when every cycle of your processor was yours and didn't flash up advertisements or reminders to get Genuine Annoyances.

In fact every instruction had a cycle count - Woe betide the programmer who did not program efficienctly as well as effectively.

0
0
Thumb Up

Over here

And ICT's first TTL logic integrated circuit designs (190xA series) used ICs made by...Texas Instruments!

0
0
Thumb Up

Nible and the likes...

I only had brief encounters with the 1401 in the "House for the History of IBM Data Processing", and it is something I can only recomend to anyone who is in the area (and is a bit of a nerd).

The two gentelmen who will guide you around are friendly and give some people the opportunety to "use" the machines.

It is absolutely fascinating how it all began and how Herman Hollerith started his enterprise.

Beware though... as far as I can remember, you have to let them know that you are comming, because they are not just open to the public as a normal museum. These people are volunteers and do this "work" and restauration in their owm time.

0
0
Thumb Up

Happy days

Great article, Must go to the Museum soon to see a 1401 working again, maybe I'll get a chance to fix it if it breaks. I was a IBM customer engineer in the 60's and we looked after 1401's and other stuff (1440,60 and1620's as well as accounting machines) in the 60's in central London, then on to System/360. What a wonderful time it was. PS yes I remember ICL wavy line printers . You've made my day..

0
0
Anonymous Coward

Erm...

>> With his SPACE machine, Underwood remembers, you could calculate the powers of 2 with a mere nine instructions. "Today, you'd need tens of thousands," he says. "That's terrible." ®

Tens of thousands of instructions to produce the powers of two? I know this guy probably hasn't touched a computer in a few decades so I'll give him a break, but on a typical x86 SAL is one instruction. Add a REP if you want to print all of them (as far as you can represent in one register, anyway).

0
0
Silver badge

Colossal Intellectual Space Race ...... Post Modern Version/UNModeRated? :-)

I would invite all who have read Cade's Mountain View "Sputnik, spaghetti, and the IBM SPACE machine" article, revealing the Past, to consider if they would recognise a Spooky Parallel Similarity in a Future Singularity Program, BetaTesting Crack Code Triggers which Crash Corrupted and Easily Compromised Operating Systems .......Collapsing Towers of Failing Flawed Facilities .... which may or may not appear on El Reg as "Who tossed in that grenade and where's the fragging pin?.. BOOM! ... Oops, too late, never mind." .... Posted Tuesday 17th November 2009 06:07 GMT ... http://theregister.co.uk/2009/11/16/quantum_processor/comments.

0
0
WTF?

What?

"With his SPACE machine, Underwood remembers, you could calculate the powers of 2 with a mere nine instructions. "Today, you'd need tens of thousands," he says. "That's terrible.""

What- has someone removed Left Shift instruction from modern processors?

0
0
Thumb Up

Dream job

"We were supposed to figure out what was coming ten years out," Underwood says. "But what we really did was screw around - and pretend we knew what we were talking about."

Where do I apply?

0
0
Boffin

@AC 10:28

As is usual for people who work outside of the embedded world you have forgotten that, although your code to calculate the powers of 2 is brief, you will need many thousands of instructions in addition to produce an executable that will work with any current OS.

0
0
Grenade

nine instructions?

Methinks he means 9 instructions to calculate powers of 2, and 100,000 to bring up the GUI, read the registry and present the data in the font and local numerical convention of your choice ;-( Or read the text file into WordPad and have it present the characters.... ;-( or... or...

Point is there is no commonly used IO device that lets you present ASCII to a human in under 100,000 instructions.

0
0
Anonymous Coward

Re: Erm...

I'm sure the quote about the powers of two was taken out of context (my guess would be that it meant something to do with factoring), but I'm still curious to know what he actually meant.

0
0
N2

Amazing

Respect to those who restored a computer thats nearly the same age as me!

0
0
Boffin

nine instructions?

re hugo @ 11:38

I remember that, as part of the IBM 1401/1410/7010 CE course, one was required to write, in machine code, a program that:

read a very simply encrypted punched card

decrypted it

printed the result on the line printer.

Then manually key that prog into storage and execute it.

Took considerably fewer than 100,000 instructions

(the simple encryption was just offsetting the alphanumerics by 1 - B on the card meant A, C meant B, and so on)

0
0
Gold badge
Happy

Best quote...

"The 1401 made a huge contribution to the 360. It paid for it."

I'm impressed. The 360's gestation was so legendarily expensive that the *book* about it was sufficiently historic to be republished in a 25th anniversary edition.

0
0

Punch Cards

What my mother thinks of punch cards having worked on a punch card fead accounts computer is not prinerbal here. Needless to say having spent many a day feeding clearing and then refeeding punch cards, she still thinks compuer is a four letter word. Sorry but I don't know which system she would have been useing.

0
0
Pint

Oh Punch Cards, you bring back memories...

Nerd that I am, as an undergraduate at university I used to write love notes to my (then) girlfirend on 80 column punch cards from the IBM 360 housed in the center of the nerd universe <the Computer Sciences building, of course>.

0
0
Happy

Legacy software.

Although i never used it, Autocoder was the first language i ever learned. I didn't start writing "paid for" code till the end of the 60's when i worked for a company that sold time on a Dartmouth Basic system.

In the early 80's i was on a plane to COMDEX and sat next to a guy reading fanfold. I asked him what language he wrote and he told me he was maintaining the back office system for a large insurance company. It ran on an IBM 370, which emulated a 360 emulating a 1401 running Autocoder!

For all i know, there may still be autocoder in a production environment somewhere.

0
0
Silver badge
Happy

My mother used 1401s

I still have all her machine usage manuals and Autocoder manuals. It's still better than using jdeveloper.

0
0
Happy

Those were the days

The high school I attended in Detroit was blessed by donations of a 1401 for the "business" curricula and a 1620 for the scientific ones. I was in the first class to learn on the 1620 (in 1965-66) and we first learned in SPS (Symbolic Programming System), a language very much like Assembler. Then on to FORTRAN II.

Weird but wonderful in there was no fixed "word" size, just a string of digits delimited by special characters giving variable precision arithmetic at the cost of CPU cycles. Also didn't have a hardware adder allowing arithmetic in anything from base0 to base36. Was thus called the CADET: Can't Add, Doesen't Even Try.

We punched our own cards for program and data input and used either the console typewriter (literally an IBM electric typewriter controled by the CPU) or the high-speed card punch followed by the interpreting printer for output.

Now that was Real Programming! :-)

1
0
Silver badge

Bah!

Whippersnappers! At President Kennedy Comprehensive School, a team headed up by one Mr Margolis, last seen at the OU, rebuilt one of these from scrap parts.

Oh the long winter evenings, tracing logic states on ten-foot long circuit diagrams and matching them to voltmeter readings on the bay interconnections.

When we finally got it to come up it was about half as clever as a Sinclair Executive calculator that kept getting its sums wrong (intermitent fault in the "C" register meant that 2+2=3 about one time in four).

Then another teacher took over the Maths department and made us break it up because he was a vengeful git with the mental age of a fourth former and an abiding hatered of anyone favoured by Mr M.

Git

0
0
Boffin

@hugo

"Point is there is no commonly used IO device that lets you present ASCII to a human in under 100,000 instructions."

Ascii ? I doubt it - Ebcdic more likely. I will check my chronology, although IBM invented them both, I think Ebcdic came first. And the instruction set had several packed to zone conversions, at least in 360.

0
0
Anonymous Coward

@ Herbert ... EBCDIC

...The 1401 was a 6-bit machine: pre-EBCDIC. Although there was an existing 6-bit `standard´ code (FIELDATA), it was primarily a military standard. My employer (the one with the wavy printers) used Fieldata heavily because they were heavily into US military contracts.

Instead, IBM used BCD coding. The position of the numerics in the BCD code space was more convenient for a decimal arithmetic machine like the 1401. A one is 01 (octal) in BCD but 61 in Fieldata, nine is BCD 11 but Fieldata 71. Both coding schemes had upper case alphabetics only, plus numerics and punctuation.

By the way, ASCII was developed by a standards committee and, far from inventing it, IBM resolutely ignored the standard for many years.

0
0

Powers of 2 calculation on decimal machine (BCD)

On a decimal/BCD machine you aren't working with binary shifts. The 1401 had a multiply so they could have implemented it with a loop though there might have been some trick to it. On binary machines, powers of 10 might have an underlying multiply by 10 that was done by load orignal value, shiftLeft2, add original value, shiftLeft1

BTW When implementing binary floating point, this was my preferred method for squeezing out succesive decimal digits for printing from a binary fraction.

EBCDIC came along five years later with the 360.

0
0
FAIL

Oh God, not the old 'back in the good old days' crap

Arr, when Oi were a lad, you'm didn't need none of yer fancy graphical user interfaces to calculate the powers of two. Aye, yer could do it in nine instructions. Now it's millions! Bloomin' madness, Oi tells yer.

Well that's us whippersnappers told then. Sure the machine might have weighed as much as a bus and cost half a million bucks, but you you could...drum roll...do the sort of wanky thing you learn how to do in first year Microprocessors Lab with a Z80 and a breadboard. Of course code was more efficient. The hardware was shite. To say you need hundreds of thousands of instructions to calculate the powers of two is ignoring the fact that those hundreds of thousands of instructions can also be used to display the results from computations that aren't so ball-achingly pointless.

0
0
Thumb Up

Dont forget the other fun thing

Loading the object deck that made those joke print outs of out of "X"'s and "O"'s that would come crackling out of those printers a line at a time, to reveal.. an image of a naked woman on a bar stool in all the greenbar glory! Those were the days!

0
0

1401 Nostalgia

You did a great job of summing up the essence of the 1401 and the speeches of Chuck Branscomb and Fran Underwood. You cannot imagine the thrills we founders had attending the events celebrating the 50th Anniversary as we recalled the period leading up to the announcement in 1959. It was not only the technology, but also the camaraderie and teamwork that made this machine and its introduction to the world so memorable.

Jan Swanson Barris, Lead Writer IBM 1401 Reference Manual.

0
0
This topic is closed for new posts.

Forums