The history of computers

Calculation aids

This is it, the first stop. The beginning of computing. Why did people begin building calculation machines in the first place, you might ask? Well, mostly they were just a bit lazy really. Here’s the oldest example we have – though not an original. It’s a cast of a Roman abacus, dating back to around 300 B.C. As you can see, it’s made of bronze. The little beads are counters, which you can use to make calculations using Roman numerals.

Moving on, here we have a reconstructed tally counter from the middle ages. In fact, this is where our phrase “tally up” comes from. A tally is both a counting aid and a way to keep tabs on something. Imagine I’m at an inn, ordering a beer. I don’t have any money with me, so the innkeeper adds it to my tab on the tally board. So this was carved in? Exactly. You can see the notches right here – nine of them in all. So that’s nine beers tracked right there on the tally board.

If you look down to this green one here, that’s medieval as well. It’s known as reckoning by the lines, as described in a book by Adam Riese. Riese was an instructor in mathematics at the University of Nuremberg, which was in Altdorf at that time. In his book, Riese explained how to make calculations based on the lines of a calculating board, comparing it to calculations with Arabic numerals. This was actually a huge problem for the fervent cleric. After all, numeric calculation is an Arabic, non-Christian invention. And that makes it a bad thing. Sounds familiar, doesn’t it? Can you explain how it works?

It’s just adding and subtracting together really. An arithmetic aid. It’s good if you know how to do it, but it’s a lot more practical using numbers.

Calculators

Good, so that’s the arithmetic aids covered. What you see now are proper calculators. This calculator designed by Schickard dates from 1623. Schickard was friends with the astronomer Keppler, who worked on calculating planetary orbits. Schickard thought it would be a nice idea to help his friend out and give him something he could use to help him calculate his planetary orbits. That’s when he had the idea of this calculating machine, which is capable of performing the four basic arithmetic operations. Schickard told Keppler about his idea in a series of letters. But at some point early on in the archiving process, the sketches were separated from his letters, and the connection was lost. Lost, that is, until a historian turned up at a mathematician’s conference and presented the material, re-establishing the link and confirming that it all belonged together, as the blueprint for Schickard’s calculating machine. Sometime later,a certain Professor Freytag von Löringhoff took it upon himself to attempt to reconstruct the machine – which he did. We have one of these reconstructions (a small number of which were made and sold) in this very collection. It’s actually a great calculator.

Crank computing engines

I remember arithmetic rulers and slide rules from my father’s time. He was an electrical engineer and, once he’d set up his logarithms and all the rest, they were very good calculators. I didn’t have to learn it at school though, thank God. I never fully understood quite how it worked either. It’s something you have to practice and keep at it to be able to use it effectively. These here, though, these are real computers, you might say. These sorts of machines are not just for private individuals, but often for companies, too. These are the kinds of machines you might expect to crop up in bookkeeping departments, for instance. Just look at the type of keyboards, which vary subtly depending on the application. As far as the mechanical principle is concerned, there are only two possibilities: either the stepped drum or the pinwheel. These are the mechanical mechanisms used to carry the tens and operate the machine. Both principles were invented – and manufactured – by Leibnitz a few hundred years ago. And both were used in such machines well into the 1950s and 60s. It works, so why reinvent the wheel? It’s hard to say which principle was better. That depended on who designed and manufactured the machine. Both principles work perfectly well, and there’s no difference on the outside as far as the user is concerned. Just a box with keys and a crank. The crank is the real heart of the design, though, because this is what you use to enter the numbers into the calculator and perform the operations. Without the crank, the machine won’t do anything.

Computing machines with electric motor

Going one showcase along, you can see what is effectively the second generation. The same machines, the same mechanism, but with an electric motor added on. These are very noisy machines, rattling and clattering contraptions. And just imagine that there used to be whole companies who did nothing but produce invoices or do bookkeeping – offices in which you would have 20-30 people, mostly women, doing nothing but typing away on these machines. Back then, the working day was 8-10 hours long, and 6 days a week not 5. People must have gone mad sitting and working on those things all day. I can well imagine that these rooms full of calculating machines were at times noisier than any metalworking workshop, with no relief from the constant clattering.

Machina Curta from Contina AG

One of the most beautiful objects in the collection is the Machina Curta. It’s quite a sad and, at times, wretched story. It was invented by Kurt Herzstark, who built this little calculator to fit in your bag. I can set the numbers here on the side. And on the top, there is a crank. All in this neat little cylinder. Then the Nazis arrested Herzstark and sent him to a concentration camp. He was a Jew, or a half-Jew I think. Something like that. He only survived because he was discovered by one of his competitors at the time. The competitor pointed out that they needed him, and Herzstark was transferred to a design department, from where the machine could contribute to the Führer’s final victory. Nothing came of it, but after the war Herzstark was able to complete his machine. The patent was bought up by a company in Lichtenstein, which, by all accounts, took him for a ride quite outrageously.

Relay computer

The next step was to try out something that might be better than the vacuum tube – which is when inventors started experimenting with the relay. Initially, relays were extremely expensive, as is always the case with a new technology. Small batch sizes that cost a lot, too. As a result, the first relay computers were more expensive than the vacuum tube computers. However, improvements had been made. The relay computers were faster, consumed significantly less energy, didn’t heat up as much, and took up less space.

How does the relay work?

Ultimately, a relay is a switch, too – only with a magnetic coil to operate the switch. When the current is on, the coil attracts the magnet and closes the circuit. And when the power goes off, it opens it again. The logic is the same as ever: 1 and 0, on and off, and I use that to build my operations.

So, they’re probably even heavier, right?

Yes, a little. But we’re talking about huge blocks of a machine in any case, so it doesn’t make much of a difference. On the other hand, with a relay I can fit double the capacity into the casing.

Vacuum tubes

Electric systems work best in tandem with a binary setup. Binary translates everything into a zero or a one, a setup I can recreate extremely effectively in an electric system by using “on” and “off.” If there is a current flowing, I have a one, if not, I have a zero. This led to the first trials with binary technology involving vacuum tubes. In essence, the vacuum tube is nothing more than a switch, an electric switch I can use to switch another connection on and off. This is the principle engineers used to begin building the first computers – machines capable of executing these binary operations. Tubes had several big disadvantages, however: they were monstrously expensive, time-consuming to produce, energy-inefficient and prone to overheating. Oh, and they burn through quickly, too. So availability wasn’t exactly the best. On the other hand, until the 1960s we didn’t have any other technology. You have to use what you have.

Transistors

The next development to emerge in the 1960s was the introduction of transistors. Now we’ve moved on to the next generation of computers. Actually, though, a transistor is no different from a vacuum tube or a relay. And ultimately, it fulfills the same function as a vacuum tube. It’s just a small metal component with three little legs. I use the one to switch the other two and turn the current on and off. Transistors do have huge advantages over the relay and vacuum tube, however: They are far smaller, use even less energy, and switch much more rapidly. Switch time is of course lost time, and with a transistor you can eliminate it completely. With the transistor, the switching is all done physically within the component using magnetization. As I said: faster and less energy, though initially more expensive. However, the price soon came down as Asian manufacturers began to produce them in larger and larger quantities, and you can simply fit a lot more into the space. Even so, these circuit boards were still wasting a lot of space. They were bulky, and it was clear that there were improvements to be made.

Core memory

While these developments were underway, we also saw the introduction of core memory. The thing about these computers was that they needed somewhere to store their information. What we would refer to as RAM. In the first instances of core memory, you could still see each and every bit. Every time these tiny circuits crossed, there would be a small ferrite core. This core would then store information through magnetization. When both circuits are conducting current, the core is magnetized. When the current stops flowing, the core retains its magnetization for a time – allowing it to store that information. The only problem with core memory is that, when I want to read it, I erase the memory at the same time. So as soon as I read what’s in it, it’s gone. That’s completely different from the RAM we have today. It’s a simple principle: cross conductors at the point at which you want to store your information; apply current to both circuits: magnetize the core; and store your information via the encoding of the cores.

Core memory Zuse

Zuse also put core memory into his Z 23. To do this, Zuse needed to get together a group of professionals capable of building the core memory. It was painstaking work and had to be spot on. If there was just one mistake, the core memory would be unusable – a huge waste of money.

So it was all done by hand?

In the first instance, everything was done by hand. Zuse subsequently assembled a group of workers suited to the work, primarily embroiderers and seamstresses. These were the people who had the eye and steadiness of hand to perform this painstaking work. Later on, they were made mechanically, or using computers as they became more and more tightly packed. But to start with, it was all done by hand.

Zilog Z80

For the next generation, we have to fast forward to the 1970s and 80s. We’re still using transistors, but primarily integrated circuits. The best thing about integrated circuits is the compacting, specifically of the legs and housing. The innards can be made much smaller, and a small IC can then fit huge quantities of transistors. That way, a CPU, which is just an IC if you think about it, can fit millions of transistors. These new developments called for computer technology – you couldn’t just do it by hand anymore. What you have to remember is that you could build the first generation of tube and relay computers by hand, placing wires and components manually on what was familiarly known as a breadboard, based on a hand-drawn blueprint.  But as for all the generations of computers that came afterwards, that required a computer. I find that a bit paradoxical: Why do I have to build the next generation of computers using a computer that is older. And worse. Does it work? What I’m driving at is the fact that we are no longer able to build computers without computers. If we were ever to run out of electricity, we’d be back in the Stone Age. In technology terms, we’d be back to the 1930s or 40s due to the fact that we can hardly use any of today’s technologies without electrical power.

First chips and motherboards

Here we have the generation of chips we use to this day. Their predecessors in any case. The 8080 CPU, the Z80. We’re in the 80s and 90s at this point. The Intel 1, 2, 3 and so on. The fascinating thing, I think, is that nothing has really changed in the basic construction of the motherboard, that is to say, the computer itself, where everything is installed. Nowadays, it’s simply a case of making the motherboards more powerful. I have the same space, but I can fit a lot more in. These days, we even have gaming PCs in which the graphics card has more processing power than the PC itself. If I want to make a computer faster, I need to free up the processor from burdensome tasks. How do I do that? By adding a frontend – between the user and the CPU, for instance – that is responsible for regulating that alone. That way, the CPU can focus on its core processing tasks, and gets everything nicely prepared from the frontend, saving crucial processing time.

Drum memory

These here are examples of core memory. In principle, all core memory is the same. It’s just got more powerful over time. These examples here were already machine manufactured.

And they still work using binary code?

Still 100% binary. Next door you can see Zuse’s drum memory – a hard disk shaped like a drum. You can see that the heads are rigid and have a fixed path. The way it works is you leave the computer to do calculations in a continuous loop where you know you’re going to get a specific result out; then, you record the result on the drum, from where you can read it out again. After that, you have to check that you got the right result. If not, you turn the read head until you have the right one. The only thing is, if you turn once too much, contact is made and you get a big scratch. In your standard hard disk, this would render the hard disk absolutely unreadable, and it would be fit for the trash. With drum memory, however, the whole thing is so spaced out that, even if you make a scratch in the magnetic surface, there is still enough material left over to the left and right to read the information.

Removable hard disks

Moving on, and we have hard disks as would be familiar to us today. There’s a whole selection here, from the clumsy bricks of the early days to the compact modern design. There at the bottom you can see an early version of the removable hard drive. These things were built into cabinets in pairs, and could be pulled out using a handle – if you had the strength. These things weigh around 25 kilos. But they are removable, nevertheless.

Removable hard disks and tape drives

These are the removable hard disks – looking rather like a stack of pancakes here. And next to them you have tape drives. The big mainframe computers common in the 1970sand 80s would have both types of data storage, both hard disks and magnetic tape drives. Each had their specific advantages and disadvantages. Hard disks have the advantage of lightning-fast access speeds, which makes them the best option if  I need to access my data fast. This in turn makes the computer faster, since it can access the information faster. Reading and writing is a speedy process. On the down side, hard disks were relatively expensive and didn’t have much storage capacity. In the case of tape drives, it’s exactly the opposite. Tapes are relatively cheap, and have lots of storage capacity, but access times are significantly longer. It takes almost four minutes to wind through a tape from start to finish. So if I’m at the beginning of the tape and looking for something right at the end, that’s four minutes before I get to my data. That’s why computers back then had both systems. They didn’t have any better option. That had to wait until hard disks became cheap enough that it was affordable to use hard disks alone.

Punch cards

The interesting thing about punch cards is that they weren’t invented specifically for the computer. In fact, punch cards originated in the textile industry, where they were used to control the first mechanized looms in 18th-century France. Until then, whenever you wanted to weave a specific pattern, you had to set the loom up afresh, and the machine would then weave your pattern. If you wanted to change to a new pattern, you had to set up the loom all over again. That could take a day or even two – time during which the factory wasn’t making any money. That’s when you began to see the construction of mechanical looms that could be programmed to weave different patterns using a series of punch cards: the holes in the cards automatically determined the pattern of threads in the fabric. In other words, I could switch my production mode relatively quickly. All that the first computer pioneers did was to copy this data storage principle, which relies on encoding information, and apply it to computing technology. You can’t really say that anyone in particular invented a new computer storage medium.

The first computer programs consisted on instructions stored on a batch of punch cards – hence the term “batch processing.”

And how exactly did computers read the information?

The punch cards would be placed in a reading device that could detect the presence of holes using a light gate. Wherever there is light present, the computer knows to equate the point with a 1 in the binary system.

Targeting computers

Here we have some analog computers. Generally, analog computers are not universal computers, but rather built for a specific application. Take this lovely brick of a thing here, for instance. This is actually a ballistic computer built for the military. Imagine I’m cruising in my ship from point A to point B, and I want to shoot my guns at another ship travelling in completely a different direction. How do I point my weapon to ensure it meets its target? That’s what this computer was designed to calculate.

It’s got to be fast then?

Indeed. The advantage of this computer is that it was built specifically for the purpose. It’s pure electronics. With digital, you have on and off, one and zero. In this case, you have fixed parameters. Analog machines measure. I have my trajectories. I have my angles. I enter the figures, and the formula stored in the machine does the rest, calculating the end result for me.

Programmable computer

Here we see the beginning of miniaturization. By the 1970s, technology was beginning to advance to the point where you could make things smaller. This was the dawn of desktop computers: computers you could fit on your desk. This was a complete novelty at the time. Now, I no longer needed a dedicated space with its own power supply to place my computer. Instead, I had a computer that fitted on a standard desk. What a huge step forwards! One of these manufacturers of desktop computers was Diehl. Diehl was known as a weapons manufacturer – but also manufactured desktop computers for a while. It seems someone had a lot of khaki green paint left over after the war, because Diehl computers are all this funny shade of green. It’s a little peculiarity in Germany’s history with computers and computing technology. After the war, many companies were left without anything to produce. After all, they could hardly keep on manufacturing military technology – that was strictly forbidden by the Allied victors. For decades, there was not a single German company manufacturing weapons. So you had companies specializing in precision mechanics and electronics that could no longer produce their old products. Naturally, they needed to find a new application – and that’s where computing technology enters the picture.

Calculator watch

(laughs) Yes. That’s what all the nerds used to have at school. I had one too.

As a calculator during math exams?

To be honest, they were much too impractical to be any use. But of course, we had to have them. They were pretty tricky to operate as I recall. You had to use a ballpoint to press the buttons until eventually they stopped working.

Were they expensive?

I remember us having to put together a good bit of pocket money.

It’s the Apple Watch of its time!

Haha, yes. Extremely nerdy!

Microcomputer

In the 1980s, microcomputers entered onto the scene. Initially, this was something reserved for the real enthusiasts. Now, for the first time, your everyday consumer could buy educational or recreational kits that would allow them to build their own computer. Imagine the electronic version of a Meccano construction set – still extremely nerdy. These gadgets didn’t really have any use as tools for home working. But, for the first time, private individuals could go out and buy their own computer. The 8-bit micro got steadily smaller and smaller.

They were just calculators then?

The big difference with these machines as opposed to the generation of   computers before them is that you could program them as you wanted. In other words, the user could determine the application. However, the problem remained that programs were lost as soon as you switched off the computer.

Home computer IBM PS/2-30

As the 80s progressed, computers slowly became an accepted part of modern life. The big breakthrough came in 1982, which was when the C64 was first showcased at a trade fair in the U.S. This was the first home computer. It is still affectionately referred to as the breadbin due to its distinctive shape. Commodore actually came up with a really clever idea for the C64. The machines were originally designed as entertainment systems. Nobody at the time ever thought that people might one day need a computer at home to keep track of their accounts or for their kids to do their homework on. No, the sole intention was to provide fun and entertainment. The Commodore C64 was a ready-to-go machine. You didn’t need to buy a monitor – you could just hook up granny’s old TV set. And in the compact box, you had these storage modules, known as data sets. And off you went.

Atari computers

Atari computers were very popular with a specific community of users, namely anyone who was involved in music or had an interest for it. This was because the Atari was the only home computer that came with a built-in MIDI interface. By plugging a piano keyboard into the MIDI port, users could experiment with different sounds on the computer, and play them back via the keyboard. It’s where electronic music began, really, allowing young people to create their own electronic sounds. For a very long time, Atari was the go-to for anyone wanting to make electronic music.

The raver’s tool then?

Something like that, yes.

The first PCs

After home computers, you had personal computers. These computers were designed to help with work tasks. There was also a drive to develop compact systems. I say compact rather than portable, because these systems really weren’t portable. On top of that, there was no battery, since we didn’t have the technology to integrate a battery fit for the purpose. Just think of the first mobile phones, real bricks that you had to lug around like a suitcase.              

Looking below, you can see a typical example of the personal computer, the IBM 2. I recently found out why computers of this era have such a weird color. The thing is, German industry had regulations for which colors were acceptable for a work machine. And that’s why they all look so ugly. Apple was the only exception: they didn’t care about those rules. Apple was the first to consider form alongside function.

CDC 3300

Here we have old and new side by side. That over there is the FAU’s present computing center. And these here are relics from the old computing center. This here, for instance, is a CDC 3300. This was the center’s first mainframe computer. As you can see, it consists of chunky hardware installed in huge cabinets. That’s why we could only preserve a portion. If you look at the cabinet containing the core memory, the equivalent of today’s RAM chips, you’ll notice what looks like a messy jumble of wires. There’s a technical rationale to the configuration, though: If you were to make the wires run parallel to each other, the electromagnetic fields in the cables would interfere with one another. In the worst-case scenario, nothing would get through, or the data would be corrupted. The point of this mess of wiring is thus to cancel out the fields and make sure they don’t disrupt one another.

TR445

This here is a TR445, as once operated here in this computing center. I really like the spaceship feel of the design – it has something of Orion about it, I find. These pictures show this very room, and how the operators would have been stationed.

Zuse Z23

Here we have a computer in its classic configuration. We have the console, that’s the input. Then over there we have the teleprinter, that’s the output. This here is the CPU, the main processor, which also contains the core memory, or RAM. And over there is the drum memory. (loud rattling). And see here, that’s the hard disk. As you can hear, it’s pretty loud. This is a computer power-up in the most literal sense. You have a rotating drum that needs to achieve a certain rate of revolutions to spin within the magnetic field without making contact – and that can take it a little while. Scattered around you can see the original schematics. Without them, we’d be in trouble. Sometimes we have to do a bit of reverse engineering. We have the machine, the schematics, and everything else, but you have to be able to interpret it. After all, we didn’t build the machine. What you have to do is approach the problem in reverse to pinpoint where the errors might be.

Edwin Aures, Zuse expert: The architecture of this machine is unique to this machine – a machine of which there are only some 150 examples in existence. This particular architecture was never continued. These are special machines.