IBM PC/XT vs Compaq Portable (1983)

Introduced March 1983

The launch of the original IBM PC in 1981 cause a fundamental shift in the desktop computing market. Out of the apparent chaos of a myriad of (mostly) startup companies offering business computers came the biggest player in the market – IBM. The original PC wasn’t the best, most innovative or cheapest product on the market by a long shot, but IBM was a serious player that business had heard of and it was an immediate sales success.

The original IBM PC had some significant shortcomings, in particular there was no hard disk and the limited expansion slots filled up very quickly. The upgraded PC/XT added a 10MB hard drive as standard, plus three extra slots, more RAM (up to 640Kb) and ROM and it upgraded the storage of the floppy drive to a maximum of 360Kb.

It wasn’t a huge upgrade over the original, but it addressed the shortcomings of the previous model well. It was probably the computer that IBM should have launched to begin with, but at over $7500 at launch, the PC/XT was really expensive.

IBM PC/XT

Uniquely for IBM, the architecture of the PC was quite easy to copy. Buying in industry-standard components such as the Intel 8088 processor and making detailed hardware specifications available made it possible for other companies to make PC clones that could be better and cheaper at the same time, and Microsoft could sell you the same operating system – MS-DOS – that the PC ran to ensure compatibility. First out of the door was the Columbia MPC 1600, but more followed.

At about the same time, another trend for “luggable” all-in-one computers was starting, with the CP/M-based Kaypro II being a popular example. Texas-based Compaq Computer Corporation combined both a PC-compatible computer in the convenient form factor of a transportable machine to create the Compaq Portable, their first product.

“Portable” was a stretch, at 13 kilograms or 28 pounds it was not an easy thing to carry. Nonetheless, it could be moved easily without unplugging a vast number of cables and components. With the keyboard clipped into place, the Compaq Portable could fit into the overhead luggage compartments on a plane, or be easily placed into the boot of a car.

Compaq improved on the PC’s architecture in their own way, broadly similar to the PC/XT. There was no hard disk as standard, but users commonly added one. The Portable was also cheaper to get started with than the PC/XT. Despite being something of a niche product, the Compaq Portable sold tens of thousands of units in its first year, and it made Compaq Computer Corporation a very successful rival to IBM in the PC market.

Compaq Portable

IBM and Compaq duked it out in the market until the mid-2000s, with IBM eventually selling off its PC business to Lenovo and Compaq merging with HP.

Both machines were the ancestors of most personal computers in use today, crucially demonstrating that the PC platform could evolve over time rather than having to be completely replaced with a new model every couple of years. And although the PC/XT and Portable were not the first PCs, and now very much obsolete, they were highly significant in developing the market we see today.

Image credits:
Dmitry Brant via Wikimedia Commons - CC BY-SA 4.0
ctgreybeard via Flickr - CC BY-SA 2.0

IBM Personal Computer Model 5150 (1981)

Introduced August 1981

Four years on from the launch of the holy trinity of the Apple II, Tandy TRS-80 and Commodore PET there was a rapidly growing (but fragmented) multi-million dollar market worldwide. Although rival micros tended to be incompatible, business systems showed a growing standardisation around the CP/M operating system and the S-100 expansion bus. Home machines had wildly different hardware and software, but tended to be based around either the MOS 6502 or Zilog Z80 CPUs.

IBM Model 5150 and 5152 printer

But in August 1981 came a paradigm shift, thanks to IBM. IBM seemed an unlikely player in the microcomputer market, specialising in powerful but incredibly expensive mainframes and whose initial microcomputer systems were also blistering pricey. It took IBM at least five years to develop a product, which was much slower than the microcomputer market was moving. IBM seemed old-fashioned in a market that was mostly dominated by younger and more agile competitors.

IBM could sense the way the wind was blowing, however. Cheap but versatile micros were finding their way into IBM customer sites while at the same time the market for big iron computing was faltering. IBM wanted a slice of the micro market, while at the same time it was aware that its traditional business processes would not be able to compete.

In a moment of enlightenment, IBM took one look at its internal rulebook and tore it up. Their entrance into microcomputers would follow a completely different path. Dubbed “Project Chess” by IBM, the development work attracts many top-flight IBM engineers to work on this new computer in complete secrecy. The result was the IBM Model 5150 - best known as the IBM Personal Computer or simply the IBM PC.

Instead of basing the PC around an IBM CPU, an Intel 8088 was chosen – as seen in the IBM Datamaster which was being developed at the same time. The PC also took a variant of the Datamaster’s keyboard and expansion slots, but then developed features all of its own. Output was either crisp text via a Monochrome Display Adapter (MDA) card to an existing model of IBM monitor, or to an compatible colour monitor with a Colour Graphics Adapter (CGA) card.

Although the PC could run a version of CP/M, the primary operating system was PC-DOS which was sourced from Microsoft. Quite how this choice of OS was made is now the stuff of legends. Initially IBM approached Digital Research (DR), the makers of CP/M, to provide a software platform for the PC. Although CP/M was designed to run on the Z80, an Intel version had been developed as well. Legend says that the boss of DR – Gary Kildall – was out flying his private plane when IBM turned up at the office unannounced, although the truth probably that DR and IBM couldn’t agree on a licensing structure. IBM then approached Microsoft and asked them to provide an OS. Microsoft didn’t actually make operating systems – their main business was BASIC – but a nearby company called Seattle Computer Products had an OS called QDOS that would run on the 8088. Microsoft bought the rights to QDOS, renamed it MS-DOS and then licensed it to IBM as PC-DOS while retaining the rights to sell MS-DOS themselves.

IBM PC with neatly-labelled floppies

Compared to PCs of even just a few years later, the model 5150 was pretty limited. The 8088 was a cheaper and more readily available version of the 16-bit 8086, but the 8088 only had an 8-bit external bus. RAM was theoretically expandable to 640Kb which was substantially more than the competition, but typical configurations topped out 256Kb. Although the 5150 supported twin floppies (up to 320Kb each) the only way to support a hard disk was to use the 5161 expansion box which wasn’t available at time of launch.

The 5150 did have a cassette interface, although almost all systems were bought with floppy drives. Typical configurations would include two serial ports and a parallel port, but eventually you could add a joystick, network card, more memory and other options. The 8-bit expansion card design was physically robust, and IBM published all the specification so that third-party vendors could make their own.

IBM had a rebadged Epson MX-80 printer available as the IBM model 5152, the most popular dot-matrix printer of the time. You could add any other parallel or serial-port printer as long as your software had the drivers for it.

The use of an open architecture (where IBM described in detail the workings of the machine) plus industry standard components made this a very flexible system. Because it was well-built and designed – albeit expensive – it became a popular business computer, although realistically it was priced too high for the home market. Third-party software and hardware followed, so within a year of launch the PC could do everything any other machine could do plus much more.

It was a huge sales success, outstripping IBM’s most optimistic projections several times over. High demand meant that most initial units were sold in the US only. Production of machines for Europe officially started in 1981 when IBM launched a plant in Scotland, but grey imports existed before that. This delay gave the opportunity for rivals such as the ACT Sirius 1 to gain a foothold.

The 5150 was the direct ancestor of almost all PCs in use today (apart from Apple’s Macintosh machines). The IBM PC XT added more expansion slots and hard disk support in 1983, the IBM PC AT came in 1984 and used a much more powerful Intel 80286 CPU. It did seem at the time that IBM was onto a winner, but it didn’t take long for other companies to build compatible machines using the same architecture.

The only proprietary part of the PC was the BIOS which had to be emulated, or in some cases just ripped off from IBM. The Columbia Data Products MPC 1600 was the first true clone of the PC, launched less than a year after the 5150. Better known was the Compaq Portable, launched in 1983, which was not only 100% compatible (and used a legal BIOS) but it was transportable too. Thousands of other companies followed suit, and within a few years IBM’s control of the market was slipping.

In 1987, IBM attempted to change the direction of the PC industry with the launch of the PS/2 range which was more tightly controlled by IBM. Clone makers needed a licence to make a PS/2-type machine which had a different hardware architecture, but few bothered and instead the bulk of the market remained with machines with a direct line back to the original 5150. IBM continued in the PC business until 2005 when it sold the unit to Lenovo.

Today the 5150 commands decent prices for collectors, commanding prices of several hundred pounds for a good one, although they are much rarer in Europe than the United States (and if importing one, you need to get a voltage regulator unless you want to blow up your power supply). Of course, you can buy a direct descendant of the original PC in any computer shop which might give you a less antique experience…

Image credits:
Science Museum Group - CC BY-NC-SA 4.0
Rama & Musée Bolo - CC BY-SA 2.0 FR

IBM Selectric (1961) and IBM Datamaster (1981)

Introduced July 1961 and July 1981

Remember typewriters? You know, the obsolete technology that existed before the obsolete technology called word processors? Somewhere after people wrote stuff down by hand? No? Oh well, the IBM Selectric probably isn’t for you.

By 1961, typewriters were clunky, slow and inflexible… but businesses everywhere relied on them. IBM had a different vision of what a typewriter could be, and the Selectric was much more feature rich than most of the machines on the market at the time.

Early IBM Selectric Typewriter

One obvious different was the print head – instead of having an individual arm with each letter laid out in a complex mechanical arrangement, the Selectric had a “golf ball” print head which would rotate to find the letter you wanted. On the Selectric, the head moved from left to right rather than the paper moving from right to left. Crucially, if the operator wanted to change the font they would just stop typing and swap in a different print head.

A quite complex electromechanical arrangement made all this work, and to get the best out of the Selectric required either experience or training. But it was faster, more reliable and more flexible than traditional devices and IBM took a large share of the business market.

New versions with more features followed, although the Selectric units were incompatible with each other. Some had correcting ribbons, wordprocessing features and even local storage. Variants of the Selectric could be used as computer printers. By the time the brand was retired in 1986, IBM had sold more than 13 million Selectric devices.

20 years further on, IBM found itself on the cusp of a larger revolution. Business computers had been getting smaller, more powerful and – crucially – cheaper, which was becoming a possible threat for IBM’s large computer business.

IBM wanted its own microcomputer and had started working on creating a unit based on an Intel processor, which was a major design break for IBM who had previously used their own PALM CPUs in their machines. The results of this unconventional effort by IBM is probably not the computer that first springs to mind – the IBM PC – but instead the IBM System/23 Datamaster.

IBM Datamaster

The Datamaster used many of the same or similar elements that would be seen in the PC, including the Intel CPU, expansion bus and keyboard. Instead of the PC’s now-familiar modular design, the Datamaster was an all-in-one box (not dissimilar to the original Mac) designed to be set up by people with no technical experience. It was also IBM’s cheapest computer to date.

Unfortunately for the Datamaster, it had been stuck in development hell and took a very long time to come to market. As it was being readied for launch, the team behind it were also finalising the IBM PC which was launched the very next month. The PC had learned many lessons from the Datamaster, keeping what was good and throwing out what wasn’t. The PC changed the world, the Datamaster found modest sales in die-hard IBM shops.

The Selectric was arguably the ultimate electric typewriter, and while the Datamaster wasn’t the ultimate microcomputer it paved the way for what arguably evolved into one. Both devices are quite collectable, although the Datamaster is much rarer than the Selectric. Out of the two, the Selectric might still be of more practical use... and your children may well never have seen anything quite like a typewriter before.

Image credits:
Marcin Wichary via Flickr – CC BY 2.0
Steve Lodefink via Flickr - CC BY 2.0

IBM 7030 Stretch (1961)

Introduced May 1961

Sometimes products are released that look like they are sure-fire successes at the time, but end up in the long run as being insignificant. Sometimes products are launched that look like failures, but end up changing the world in some way. The IBM 7030 Stretch is a little of one and a little of the other.

The 7030 was IBM’s first fully transistorised computer, and at launch it was the fastest computer in the world. Projected to be priced at an eye-watering $13.5 million dollars in 1961 money (about ten times that today), this was a serious computer for serious organisations – coming in at 32 metric tons and consuming 100kW of power.

Transistor technology had been developing at a rapid rate by the start of the 1960s and IBM proposed using diffusion transistors for the new design. This was a risk move for the typically risk-averse IBM, but competition with companies such as UNIVAC was heating up. The initial goals for the 7030 was impressive – a 64-bit system capable of a processing capacity of 10 MIPs. When the technical complexities of the project began to dawn, this was dropped to 4 MIPS. When the 7030 was launched, it actually shipped with 1.2 MIPS. 

IBM 7030 Stretch

The system performance was a disappointment – even though it turned out that the 7030 was the fastest computer in the world. IBM cancelled new orders and halved the price for those who had already ordered it. In IBM’s eyes, the 7030 was a failure. Just 9 units were sold – including one secret version known as “Harvest”. There were significant internal recriminations at IBM, with plenty of finger-pointing going round and people anxious to assign blame.

But the 7030 was more of a technological success than was realised, and the innovations in hardware and software found their way into other IBM products, especially the successful IBM System/360 series which found their way into corporations everywhere. And although many of the technologies in the 7030 were soon obsolete, they all provided an important stepping-stone in the development of 1960s computing.

IBM 7030 Stretch

Two key figures in the 7030 Stretch were Gene Amdahl, a legendary designer of powerful early mainframes and Frederick Brooks who went on to write the seminal software engineering tome “The Mythical Man Month”. This book attempted to learn from the mistakes in the project management of Stretch and other projects, significantly the idea that adding more people to a late software project will only make it later. Despite being in print for more than 40 years, corporations continue to make the same mistakes that IBM did in the early 1960s.

Image credits: Don DeBold via Flickr - CC BY 2.0
[1] [2]

Hollerith Tabulating Machine (1890)

1890 Hollerith Tabulating Machine

Introduced June 1890

The United States has a census every ten years, an essential act of counting all the population of the country and working out their demographics. But as the population grew in the 19th century, so it took longer and longer to take the census and process all the information. By 1880 the whole process took eight years, nearly long enough to clash with the 1890 census. Something had to be done.

The key part to processing the census more quickly (and cheaply) was the punched card. First introduced in 1804 (more than two centuries ago!) with the semi-automated Jacquard Loom, punched cards allowed binary data to be stored and read by simple mechanical and later electro-mechanical devices. It seemed to American inventor Herman Hollerith that this could be a key part of the solution to the census problem.

For the new census, data was still collected on paper but it was then transcribed to a punched card with 12 rows and 12 columns of binary data, marked by the presence or absence of a hole. The key element was Hollerith's Tabulating Machine. An electromechanical sensor combined with a simple counting dial could then add up the data in a variety of ways, which allowed for all sorts of data analysis.

100 million cards were made, and each was processed just four times to come up with the variety of statistics that the census office wanted. It took two years off the time it took to process the data, but the lasting legacy of the tabulating machine was much deeper. For the first time it showed that automation could be used to process data on a large scale. And remember… this was 1890.

Hollerith’s business grew and in 1896 he create the Tabulating Machine Company, which then in 1911 merged with some other businesses to become the awkwardly named Computing-Tabulating-Recording Company. In 1924 this company renamed itself to International Business Machines (IBM). And a century after the Tabulating Machine’s success in the 1890 Census IBM gave us... errr… the PS/1.

Punched card from a Hollerith Machine

As it happens, the Hollerith Machine could be used for evil as well as good. The Nazi regime used demographic data collected on punched cards extensively with horrifying consequences.

Hollerith’s tabulating machine popularised the punched card, something still seen today sometimes in voting machines but which generally fell out of use in the 1980s. Today punched cards can be quite collectable, especially for more obscure systems or ones with some historical interest. However, most people who actually did use them don’t miss them at all... especially if you’ve ever dropped a big pile on the floor and have then had to sort them back into order by hand.

Image credits:
Diane Maine via Flickr – CC BY-NC 2.0
Marcin Wichary via Flickr – CC BY 2.0

IBM PS/1 (1990)

IBM PS/1 Model 2011

Introduced June 1990

By 1990 the PC-compatible marketplace had changed a lot since the launch of the original IBM PC (model 5150) in 1981. No longer just the choice of businesses, PCs had largely replaced an eclectic range of incompatible home microcomputers that had dominated the earlier 1980s. It was increasingly common to see PCs in the home, but they weren’t generally IBM PCs despite IBM inventing the platform.

IBM had tried to break into the home computing market in 1984 with the IBM PCjr, a short-lived crippled version of the PC that was a sales catastrophe. Apparently unperturbed by this, in 1990 IBM tried to break into the same market again… and they repeated many of the same mistakes they had done years earlier.

Worse still, IBM’s attempt to redefine the business PC market with the IBM PS/2 launched a few years earlier was floundering. Instead of moving the market from DOS and the old ISA hardware architecture to OS/2 and Micro Channel Architecture (MCA) it seemed that IBM just split the market between themselves and competitors such as Compaq who were improving the old platforms instead.

In 1990 IBM tried a shift in direction with the new IBM PS/1. Rather more based on traditional PC architecture than the PS/2, it was designed for home users who wanted to be able to unpack something from the box and get working in minutes Models such as the 2011 made this really easy, and when assembled they booted into a friendly screen allowing easy access to DOS, Microsoft Works on online services if they had been included.

IBM PS/1 Model 2133

A Mac-like simplicity to the hardware had some drawbacks – it wasn’t really expandable and the non-standard power arrangement where the computer was powered by the monitor (like the Amstrad PC1512) meant that you were stuck with using the IBM PS/1 monitor for ever.

The hardware was excellent though, and it wasn’t stupidly expensive (competing with Compaq on the likes of price), but consumers were not that interested. It didn’t help that IBM had to create a completely new sales channel for the things as traditional IBM dealers didn’t sell to consumers, but in the US large-deal with Sears who bundled access to Prodigy with the computers. On early models DOS was included in ROM, which made the machines very quick to boot up.

Consumers were cool about the PS/1 though, preferring other brands where they were available. IBM was still seen as a business PC, and the incompatibilities of the PS/2 range rubbed off on the PS/1 even though it was a different hardware platform. IBM stuck with the range though, making the machines more expandable and more standard in terms of hardware and software.

The range lasted until 1994 when IBM replaced the PS/1 range with the architecturally similar but more appealing IBM Aptiva range which continued until IBM’s exit from the home PC market in 2001. Today the PS/1 is an uncommon beast but it commands decent prices of about £500 to £700 or depending on model.

Image credits:
Kungfoocow369 via Wikimedia Commons – Public Domain
Science Museum, London – CC BY-NC-SA 4.0

Jerusalem Virus (1987)

Not a screenshot of the actual virus

Discovered October 1987

Thirty years ago we saw a maturing of the personal computer market, but this was also joined by the rise of the phenomenon of malware. 1987 brought not only the Stoned virus, but also the infamous Jerusalem virus… discovered in Jerusalem in October 1987.

Where Stoned wasn’t deliberately destructive (but was accidentally, due to bugs), the Jerusalem virus was. On each Friday 13th (except for 1987), the virus would try to delete any program run on the DOS PCs it infected. It was an executable file infector, adding its code to the applications themselves.

Of course, this would be a trivial thing to recover from if you had the original installation diskettes, but in those days it was extremely common for software installed on a PC to be a copy of a copy of a copy, and many people didn’t have the disks. And of course, the act of copying software itself helped the virus spread from infected PCs to other PCs via infected programs on the floppy disks.

Back in 1987 anti-virus software was also in its infancy, with products such as McAfee VirusScan being early entrants into that market. Few people had anti-virus software, and given the high levels of piracy of applications it took some years for the first wave of computer viruses to be brought under control.

Later variants of the Jerusalem virus were created, but eventually they all vanished completely. File infecting viruses do still exist these days, but are still quite rare and old DOS viruses such as Jerusalem won’t even run on Windows. However, thirty years of PC malware evolution have led to things that are much, much nastier than the Jerusalem virus.

Image credit: Gugganij via Wikimedia Commons

Stoned Virus (1987)

Stoned virus  hexcode

Created 1987

Computer viruses and other malware have been around for a long time, with early reports going back as far as 1982 with "Elk Cloner". However, one of the first really widespread viruses made an appearance thirty years ago - the Stoned virus.

Stoned first appeared in New Zealand in 1987 and spread on IBM PC compatibles via floppy disks. Now, it's quite possible that you have never used a PC with a floppy disk as they largely vanished from new PCs in the early 2000s, but on an early PC the floppy disk would be the A: drive (and if you had a second floppy, it would be the B: drive).

PCs would boot from the floppy first, even if they had a hard disk and because people would tend to leave floppy disks in the drive when they powered off the machine, this gave the virus an opportunity to spread. The PC would attempt to boot from the infected floppy and appear to fail - at which point most people just removed the floppy and pressed a key to boot from the hard disk. Unknown to them, the Stoned virus was then in memory and it would write itself to the hard disk when the machine booted up. One on the hard disk, Stoned would then try to infect all floppy disks put into the machine, and if an infected floppy was taken away and put in another PC then the process would begin again.

Ancient IBM PC of some description

The PC would sometimes display messages such as “Your PC is Now Stoned!” and “Legalize Marijuana” on boot up, and it would tend to corrupt anything other than basic 360Kb floppies. Other than that, it spread quietly.

And because the most common way to share files in those days was to swap floppy disks, spread it did. Not quickly at first, but like a zombie apocalypse eventually almost every PC in an organisation could become infected. And of course, any contemporary anti-virus product would stop it… but in those days many organisations didn’t take malware seriously or thought that security products were too expensive.

The virus continued to infect machines for years, but even though anti-virus software could stop the PC becoming infected then millions of floppy disks with it on meant that it would keep trying to come back. Eventually of course floppies fell out of fashion and then vanished altogether – and it’s quite likely that those decades-old disks have now degraded to the point of unreadability.

It wasn’t the last time that a virus used similar techniques to spread. The Conficker worm from 2008 could spread through USB devices and is still a problem nearly a decade later. Even more ancient malware such as the 2003 SQL Slammer worm still flares up from time to time. Of course, malware is not a gadget… but it seems to be an unwelcome companion to technological advances.

Image credits: Wikipedia and Luke Jones via Flickr

IBM PS/2 (1987)

Launched April 1987

In 1981 IBM launched its first mass-market microcomputer, the IBM PC. The upgraded XT followed in 1983 and in 1984 the significantly more advanced 286-based AT hit the market. Based mostly on off-the-shelf components, the IBM PC became a huge success.. but it didn't take long for upstart competitors to make computer that were either faster or cheaper than IBM's (often both).

With the PC, IBM had created a monster but they began to realize that they no longer controlled the monster. IBM's response was to try to create a new monster that it could control more easily. As a result they developed the IBM PS/2 and the OS/2 operating system. Neither succeeded.

IBM PS/2 range

There were technical limitations with the IBM PC. Essentially based on software and hardware rooted in the 1970s, IBM was looking forward to the 1990s. The idea with the PS/2 was to introduce multitasking, plug-and-play expansion cards, better memory management and of course more powerful processors. Obviously, these were things that customers wanted too - but IBM decided that customers didn't need much in the way of backwards compatibility, and customers decided that... well, they did actually... and they stayed away in droves.

It's easy to criticise IBM for wanting to push things forward even if customers were resistant, but the place IBM was in during the late 1980s was not the same place as Apple in the late 2010s. Companies such as Compaq could offer something better than the IBM PC while still being compatible, where the PS/2 wasn't. In changing direction, IBM lost what little control it still had over the PC marketplace.

The OS/2 operating system was late as well. Originally developed by both IBM and Microsoft the idea was to create a next-generation version of Windows based on the OS/2 core. However, the first really usable version of OS/2 appeared a year after the PS/2 leaving early versions of the hardware running IBM's version of MS-DOS (called PC-DOS). It took until the mid-1990s for OS/2 to become rather good with the launch of OS/2 Warp, but by that time it was competing against both Windows 95 and Windows NT.

The PS/2 soldiered on into the 1990s and it was accompanied by some desperate efforts by IBM to regain market share, such at the PC compatible PS/1, Aptiva, NetVista and ThinkCentre ranges. It took IBM nearly two decades to throw in the towel, divesting the PC business to Lenovo in 2005.

Oddly enough though, the PS/2 was hugely influential in other ways. The PS/2 introduced VGA graphics and the 15 pin D-type cable still seen today (just about), it helped popularise the 3.5" floppy drive and memory on 72-pin SIMM modules, and it created the mini-DIN connection for mice and keyboards commonly known as a PS/2 connector. All of these features ended up in the products of rivals. Not least the PS/2 range was nice to look at, giving a welcome boost to standards of industrial design.

Image credits: Raymangold22 via Wikimedia Commons

D-Kuru via Wikimedia Commons

IBM 3800 Laser Printer (1976)

IBM 3800 Printing Subsystem

Announced April 1976

Announced forty years ago this month, the IBM 3800 was the world's first commercially available laser printer. In one form or another the 3800 was in production for a decade-and-a-half and revolutionised high-volume computerised printing.

Beating rivals Xerox to market by a year, the IBM 3800 was a massive device that could typically print about 167 pages per minute. Today, even a high-end device such as the Ricoh Pro 1357 would struggle to keep up. Crucially though, the IBM 3800 used continuous stationery rather than cut-sheet paper which must have helped with the throughput. The print resolution was about 160 dpi.

It wasn't cheap - even the 1987 model cost a staggering $175,000, but then you would hook it up to an IBM mainframe which would have cost you a couple of million dollars. Strictly an enterprise computing component, the 3800 must have been a welcome relief from the deafening noise of banks of impact printers.

It wasn't until the mid 1980s that laser printing because remotely affordable on the desktop, with the breakthrough product being the original HP LaserJet. IBM span off the printing business as Lexmark in 1991, which still exists today.

Today you can pick up a new small laser printer for less than €50, with typical very high volume systems coming in at €35,000 or more.

Image credit: IBM Archives