Manchester Baby's 70th anniversary

June 2018

Selection of valvesThe Small-Scale Experimental Machine (SSEM) can lay claim to being the world’s first computer. Martin Cooper MBCS caught up with Chris Burton, from the BCS Computer Conservation Society (CCS), to find out more about the original machine and how he helped to recreate it years later, and to learn what the future holds for the SSEM.

‘So, the Baby - and it was called the Baby by the pioneers - was really called the Small-Scale Experimental Machine,’ Burton begins. ‘Immediately after the war, word was about in scientific circles. There was talk of these new-fangled computers.’

In 1946 the Electronic Numerical Integrator and Computer, or ENIAC, was announced. It was one of the very earliest electronic general-purpose computers and was used for calculating trajectory tables by the American military. It used, Burton says, 100,000w of electrical power and employed 18,000 valves. ENIAC was physically enormous, but lacked a program memory.

At much the same time, John von Neumann began to think about, and write down, ideas about how a computer could be physically articulated. A Hungarian-American, von Neumann enjoyed a career as broad as it was successful, spanning, as it did, maths, physics, computer science, statistics, and more besides.

In 1945 he described a computer architecture that consisted of elements such as a processing unit, memory and external input/output units. The concept become known as the von Neumann Architecture. It helped shape the SSEM and, by extension, your modern PC’s internal workings too. The CPU as brain, RAM as volatile memory, hard disk as permanent storage: This ubiquitous and enduring model owes it all to von Neumann.

The perfect time and place

Elsewhere computing’s essential ingredients were coming together. Beyond ENIAC and von Neumann’s work was Turing’s 1937 paper, ‘On Computable Numbers’. Add to this Bletchley Park’s code breaking work and the Colossus code breaking machine, and computer scientists of the day began to draw together a technical recipe that would change the world.

‘The Bletchley work was only known to a very small number of people,’ Burton says. ‘But computing was in the air. People felt that they could get on and build a real electronic computer that could do real calculations.’

One of the most prominent computer experts of the day was Max Newman. He was a mathematician and codebreaker whose work led to the construction of Colossus. ‘He couldn’t breathe a word… He couldn’t say “I worked on a computing machine at Bletchley”,’ Burton says. ‘But he, and his friend Turing, knew the principles of how you could build a computer and they had a crude example in Colossus. It was all in their minds and, after the war, Newman went to Manchester University, as professor of mathematics, and had an ambition to build a computer.’

Building a team

Elsewhere, two other key actors in computing’s story were seeking a new challenge. Burton says: ‘Freddie Williams and Tom Kilburn were radar men working at the Malvern labs. They were both top notch engineers who knew how to use the technology of the time. But, after the war the urgency of radar development was reduced, and they sought new fields.’

Williams was aware of the requirement for a computer memory system and chose to investigate the application of his knowledge of the properties of radar cathode ray tubes to the storage of data for computers.

‘Computers didn’t really exist at the time,’ Burton says. ‘But he set to and started building a memory system based on cathode ray tubes. And, cutting a very long story short, Williams went back to his old university, in Manchester, and became head of the Electro-technics department. We’d call it electrical engineering now. He took Kilburn with him and they continued to work on the memory project.’

Joined later by Geoff Tootill, the three were eventually successful and built a memory system that could operate at microsecond speeds and hold binary data. ‘This was the goal of many groups across Europe and the US,’ Burton says. ‘And Williams, Kilburn and Tootill did it first - in Manchester - at the end of 1947. They then built a minimal computer to demonstrate that they had built a good memory system.’

The computer they built in Manchester was, of course, the Small-Scale Experimental Machine - better known as the Manchester Baby. ‘It worked,’ Burton says, ‘as a computer on the twenty first of June 1948.’ Pausing for emphasis, he then says: ‘And that’s where it all started. That was the first time, anywhere in the world, that a program stored in an electronic memory was executed and completed.’

From small acorns

What did the program do? Thankfully documentation has survived, including one key source of insight: Geoff Tootill’s lab notebook. In this book, conservationists have found the actual code that Baby ran. The general consensus is that the first program - which consisted of 17 instructions - was designed to find the highest proper factor of any given number such as two to the power of eighteen (262,144). The program was written by Kilburn.

Freddie Williams is reported to have said: ‘Each time we put a program in and pressed the start switch the display entered a mad dance with no useful results. But, one day, shining in the expected place was the expected answer. And the world was never the same again.’

And what was it like to program that SSEM? ‘You have to think in binary,’ Burton says. ‘Intellectually it’s trivial, and loading a program is very laborious.’

The Manchester Baby's legacy
The Manchester Baby’s importance, for Burton, resides in history and in heritage. ‘If you look at steam engines’, he explains, ‘and explore how they developed, you can dig back to the days of Watt and Newcomen and other pioneers. Today’s engineers will say that those were the seminal days.’ Burton continues: ‘In engineering it’s where something important started. And this applies to the Manchester Baby.’
The Baby showed how one computer could, through running different programs, do different jobs and solve different problems. And this universality is what makes it so critically important in computing’s evolution.
Previously, individual computers were built to solve specific problems. The Baby changed this. Today, of course, as you sweep across a phone screen full of different apps, this might sound trivial. But, back in the 1940s, the concept was revolutionary.
‘It was a universal piece of hardware. It’s a universal machine and it can do anything - if its within its scope, even things it’s never seen before,’ Burton says. ‘If you look back through the generations of computing… laptops, mainframes, minicomputers… they are a bit bulky, but they do the same thing, only on a different scale. The point is though, you can’t go any earlier than the Baby. That’s why it’s important.’
So, is the Baby the world’s first computer? In Burton’s view that’s a deceptively simple question, one that belies tides of complexity, nuance, vested interests, and opinion. ‘But’, he says, ‘if you make it a simple, straightforward question: “What is the world’s first computer?” you can’t go any further than the Baby.’ ‘It was’, he says emphatically, ‘the first computer.’
Rebuilding the Manchester Baby
The Small-Scale Experimental Machine’s 50th anniversary took place in 1998. To mark the event, Chris Burton had instigated, and led, a team with an ambitious goal: to recreate the Baby. Though little of the original machine still existed, the team had access to schematics, diagrams, and notebooks all created by the Baby’s original builders.
‘Manchester University and CCS were’, he says, ‘both very supportive and, after a lot of hard work, the machine was recreated - with ICL acting as sole sponsor.’ In all, a group of six volunteer engineers did the design work and built the recreation.
And it was a success: at 11am on 21 June 1998, the SSEM replica re-ran the first ever stored memory program. The replica was donated to the Manchester Museum of Science and Industry - where it lives today and, far from being retired, it gives regular demonstrations of true vintage computing. ‘The ambition is’, Burton says, ‘to keep the replica going for another decade, though’, he concedes, ‘skills, spares and materials are becoming increasingly scarce.’

Image: Getty/rudigobbo